Butterfly Network Stock

All right. Perfect. Thank you, everyone, for joining, and I'm excited to welcome the Butterfly Network team. There is a disclaimer on the Morgan Stanley website that I encourage everybody to go check out. We have a lot of questions to get through today. So I think we're going to just get started here.

Joe, maybe I'll start with you and giving a little bit of the strategic direction that Butterfly is headed. What's your vision for what Butterfly can become? And how has that changed over the last few years, particularly as opportunities beyond the traditional point-of-care ultrasound have emerged?

Well, yes, we've [indiscernible] first of all, thank you for having us. It's really a pleasure, and I appreciate all the work that you've done.

So we've learned that we have built a very strong technology and patent portfolio to be the leader in semiconductor-based imaging. And I think that's a new concept for us. We built our business and our platform around a semiconductor-based ultrasound device that can image the entire body at low cost that can be used to image anywhere in the body, anywhere in the world. And that is our core mission.

But the concept of democratizing imaging through a semiconductor, that aperture has grown quite greatly recently. We've learned that our core technology can be leveraged for many different clinical use cases. Butterfly still is a company that's developing a large market in point of care. But by licensing out our tech, working with partners on new use cases, it brings shareholders into a much larger TAM and a much larger opportunity.

So I think what we've learned is we've solved a major -- a foundational technology problem that serves the point-of-care market and many others. So we're pulling our core technology to the center of our circle, and we're leveraging it for markets, all different types of markets, of which one of them is point-of-care ultrasound that we're developing ourselves.

Okay. Great. I think it's important for people to understand your competitive advantage and how parts of that competitive advantage actually might become even stronger as the company scales.

Yes. Well, I really think the big imaging made a mistake. Big imaging looked at semiconductor-based imaging as something that can never really go upstream, that can never get higher image quality that can never challenge a lot of the core tech and they're wrong. So we are benefiting a lot of the -- and when I talk about the part of imaging that we're focusing on, it's not the back-end image processing and all the work that could be done. That could be done by anything. It's about being dynamic on the front end on how to capture the image and how to be dynamic about that image capture.

And I think they didn't think we can get to the higher levels of imaging. They didn't think we could accomplish harmonics, which is a major threshold of moving sound easier through the body to get that higher image. Our fourth-generation chip that will be out next year will be harmonic. And so we will have proved their hypothesis wrong.

And when you think about the competitive advantage, it cost us $300 million to make our first chip. We'll launch our fourth chip in early '27 and with the best handheld imaging device on the market, not the best device, the best imaging device. So with 600 patents with hundreds of millions of dollars of know-how and supply chain, we're going to partner with the best chip fab in the world. TSMC is our partner. Starting -- standing start. 5 years for someone else, and they have to figure out how to get around 600 to 700 patents.

What I think has been fascinating is you have a strong core POCUS franchise, but then you also have this very early stage, but extremely promising and potentially larger opportunity with the Butterfly Embedded market. How do you think about investments across both parts of the business? And how are you thinking of prioritization in particular?

So Butterfly, to the credit of our founder, Dr. Rothberg, Butterfly was architected with a very modern architecture upfront. What do I mean by that? I've been in health care for 35 years, and I've run a lot of businesses in health care. And I've had devices that have software in them. And if I wanted to -- it's actually sadly true today. If I wanted to upgrade the software of my devices in hospitals, I'm sending a rep in with a USB stick, right?

Butterfly originally was architected to be completely cloud-based. So we have 150,000 devices around the world, probably 80,000 of them are active after all the upgrades. And we have -- and I know where -- or we know where every single user of the device is. We know what type of scan they're doing and their scans come into our cloud. We have 25 million images -- 25 million cines, which -- a cine is a collection of images, probably over 1 billion images in our cloud, and it's growing by 40,000 cines a day.

So we were architected to be cloud. We are architected to have iOS and Android applications. We have a unique device with a semiconductor. And then we have a SaaS-based middleware that connects all of what we do into pulling the data into the EMR, into the DICOM and into revenue cycle management.

So all of those capabilities that Butterfly built for ourselves, as we start getting into conversations with third parties, they want to license the chip and a development software to modify the chip. But as they start getting more mature and looking at commercial, they want to have the ability to stream all their images into our cloud. They want to all of a sudden now start using our middleware to capture their data.

So what we're doing with our architecture is we're taking the point-of-care ultrasound business from the center of the circle and pushing it out to the outside of the circle and saying this is a sales and marketing effort in point-of-care ultrasound. Then we're taking our cloud capability, our AI, Android capability, our data aggregation capability with health systems and our semiconductor capability and pulling it into the center of the business. And we're modularizing all of those components.

So we then develop in a single-minded way of advancing all of that. And then we provide to point of care everything it needs, and then each of our partners now, we have brain-computer interface partners. We are -- then they will turn around and say, "Well, we want this version of the chip, and we want this SDK. We want to have this application. We want to have access to this data in your cloud." And it turns into we're this technology company that services all these different capabilities, if that makes sense.

Yes. No, it does. And I do want to spend some time unpacking Butterfly Embedded in particular. The Midjourney and Merge partnerships really validate the technology and applications beyond traditional handheld ultrasound. How do you think about those partnerships? And how have they changed the way you think about the addressable market that Butterfly has a right to win in?

So both companies, Midjourney and Merge, are very sophisticated and have -- I won't say unlimited funding, but access to -- they're not capital constrained. And they have very big dreams. If they could reach their dreams without Butterfly, they would. And so we've gone through very significant negotiations. We've gone through IP. We've gone through everything. And we're delighted to have them as partners because they are literally pushing the boundaries of what's going to happen in the future.

So what we've learned is we have a business like this is real. Forget about, "oh, I'm going to let someone use my whatever and charge them whatever for it." I mean, what's valuable to them is over a decade of evolution of technology development. What's valuable to them is the -- they can stand on top of our intellectual property platform and start filing IP on top of ours. So they get all the dates from the original Butterfly of what we've innovated and it creates a foundation for their business to solely stand on our shoulders and innovate past where we are in their certain use cases.

And so those deals -- we have other deals with very sophisticated partners that we are not at liberty to talk about. But those deals helped us validate that we have a really serious foundational technology. We have a significant lead. We have intellectual property that protects us and is extremely valuable. And we have a partner that can scale to the type of volumes that these type of companies would want to scale.

I can't talk at all about what Merge is doing other than what's in their press releases, but they're going for very big TAMs. And so for us to be a part of that, for us to be an enabler for them, it's going to constantly keep our need to be at the cutting edge of what we do. So we're not like, oh, we have this 5-year lead on everyone else, and we're just going to sit back and enjoy that. We -- they're pushing us to the boundaries of data acquisition, of AI, of processor capability and speed. It's an extraordinary experience for us.

Yes. I mean, the way you've described it is it feels like the long-term ambition of Butterfly Embedded is to become an ecosystem, right, where other companies are taking your technology, building products on top of that, more of a platform model versus you all needing to develop every single end market application that's possible.

Correct. I mean we -- our P&L is full of market development activity for point-of-care ultrasound, direct sales forces, conferences, clinical work, like all the money we spend in our digital activation to get to market is a lot. But we are now bringing to shareholders so many -- so much more TAM. I think we took a stab at it and it went from like a $20 billion to a $250 billion TAM, which is nuts by just -- again, sometimes companies solve problems that other people have. And there's many examples in history of companies doing that and then monetizing that foundationally across the industry, and we're just the next one to do it.

Yes. And I mean it does beg the question, though, as you're evaluating partners, which ones are better to enter through partnerships, meaning which -- some applications are better to work with a partner on versus some where you have the capability set and you want to own the end market and the end product and the economics itself. How do you think about that decision?

So we've kind of ring-fenced what we believe the Butterfly opportunity is. And the Butterfly opportunity basically has to do with general medical imaging, where on one side, like we are, let's say, in the middle where we have a probe for every doctor and they can bring it in and out of the hospital. We are going to go upstream into the hospital over the next several years. Our fourth generation chip that comes out in the beginning of '27 will have image quality better than any other piezo handheld in the world.

So we're going to start asking the question, why would you have an analog piezo handheld if the semiconductor -- it's like why would you walk into Blockbuster and you can just download the film on Netflix? Why? And it's the same experience that digital photography had. Kodak came out with the first digital camera. 2 years later, they shut it down because there was 1 megapixel image. So it didn't matter if it was cloud. It doesn't matter if you can store and forward. It didn't matter that you didn't need film, the image quality wasn't good enough. So all those digital benefits weren't worth the fact the primary outcome was substandard. But the moment you got to equivalent and even better, all the digital capabilities now come into the purchase decision.

And that's where we're at now. We're going to turn medical imaging, core medical imaging on its head, and we're going to capture more and more. We're going to be launching our own cart in the next year, which is going to bring us into about an existing $3 billion market opportunity of existing revenue upstream. And we're also going into the home to be able to help nurses diagnose or capture images and -- where they are to keep them there and keep them out of the hospital. And so those are our core markets.

Now there are companies that have very specific use cases in women's health and others, and it would cost us so much money to go and develop that individual use case. It's almost like building another point-of-care business. But we get to participate in that business now because they're going to do that market development. So wherever there's very specific niches or very specific clinical applications that we're not going to develop ourselves and someone is there, we're going to choose to partner.

Okay. That makes sense. I mean the way you talked about the TAM opportunity growing so substantially, do you feel like what Midjourney and Merge partnerships are, are representative of the total opportunity set? Or do you think we're still really early on in just discovering what is the power of this technology?

Well, I think it's still very early. I think Merge is our fourth or fifth BCI company today and there will be more. And so we've learned that in communicating with the brain, you can put a wire in the brain, there's an amazing company called Science who's actually growing brain neurons into the chip, into the brain or you can use sound waves to communicate with the brain.

And there's a lot of vision there from dealing with chronic conditions or dealing with trying to restore function to ways of enhancing the brain and ultrasound is the least invasive way to be able to understand neural activity. And we've even had partners start imaging through the skull, which people thought was not possible with ultrasound. And if you're interested, Aleph neuro, A-L-E-P-H neuro have some images online that are some of the most extraordinary images of a living brain ever taken through the skull. But there are so many applications in intravascular, cardiac ablation. They're in HIFU on a microscale. There are so many different applications of the technology that we haven't fully discovered yet.

And then we are actually talking quite actively to our first embedded partner that's not health care. It has to do with a company that does material stress. So if -- like an airplane wing, if it's under a significant amount of stress, you can see the materials start changing form prior to a crack. And so if you're pushing an envelope of something, and there are all kinds of construction and military applications where you can sense the changes in these materials.

And so I think we actually today don't fully understand the full opportunity. We've done the best we can. And we've now, in the beginning of the year, have started to staff up the part of the business to be able to be much more proactive. Most of the relationships we have today were all inbound to Butterfly.

Yes. I mean as you explore other industries within health care, beyond health care, how does that change the business model fundamentally, the org structure, your capital allocation priorities? I assume it's kind of an evolving decision process.

Yes. I'll take that one. Joe touched on it earlier. One of the beauties of what we're doing and what we have is we've already made $300 million of investment for the core platform that we have and the 600 patents that we have. So our ability is to leverage that, and we're leveraging that into our core POCUS business as well as leveraging that into Embedded.

So the strength for us going forward will be that leverage that we have from an operating perspective as well as from the investment that we've already made. And working with our partners, we're learning as well. So it's kind of like a virtuous circle where we're taking that learnings back into our business, and we can use it across other -- our existing business plus other partnerships provided it's not an exclusive relationship like what we have with Midjourney.

Right. Okay. Well, you said Midjourney, and I did want to go a little bit deeper on that one. What does Midjourney's approach to whole body imaging tell you about ultrasound's role in health care and how it can evolve over the next decade?

Well, first of all, ultrasound is safe. You don't have to worry about a piece of metal being ripped out of your body if you didn't properly communicate to the nurse. You don't have to worry about radiation and how much radiation they'll have. It's something that can be done very frequently, weekly, biweekly, monthly.

And I think Midjourney's vision is this isn't -- we have a sick care system today. We don't have a health care system today. And what I mean by that is if you want to know what's going on in your body, you can't just go get an MRI. If you can, you're going to have to pay $5,000 and go through a bunch of hoops. But an MRI happens when you've injured yourself. It happens when a doctor thinks you're sick. It happens when you might have some type of risk factor that people want to look at. And then you have to get a prior auth from a payer, you have to get a prescription from a doctor.

So it's highly captive, highly regulated and high cost and highly necessary. I mean it's -- I'm injured, if I'm sick, I want an MRI. I'm not talking against an MRI. But there's a whole segment to health care, which is about as you're younger, learning about your body, how is your body changing? How are the behaviors that you have as a human being affecting your health, your diet, your alcohol consumption, your smoking, your stress, your lack of sleep. And we all see this with Apple Health, with Oura, with WHOOP. There's this desire to understand how we are progressing through life.

And I think Midjourney takes it to the next level, where Midjourney is going to democratize imaging. They're going to make it really low cost to be able to get a scan. And it's going to be a part of your life. You're going to be able to know how is my visceral fat. How -- I've been working out at the gym on arms, are my arms getting bigger? And these things might seem pejorative, but it's knowing about your body and knowing that behaviors that you are taking are changing your body is going to allow -- that's the only way our health care system is ever going to succeed on the back end because poor Medicaid has to take on Medicare, has to take on 65 years of bad decisions every year instead of finding a way.

And there's no constituent in health care that cares about your health. Your employer has you 3 to 5 years. They want you to have the best health possible, but they're not there for your entire life. Your payer -- how many payers have you have? How many different jobs have you had? And so the health systems just want to be able to manage you. And if you're lucky enough to live in Rochester, Minnesota for your entire life, you can be embraced by the Mayo Healthcare system and you'll have the best care ever. But the moment you move, you're starting that clock again.

So I think what Midjourney is teaching us, first of all, is ultrasound is safe, ultrasound can be an everyday part of your life and ultrasound is inexpensive. What they're also teaching us is that this is a text-to-image company. This is, let me take a little bit of data and let me extrapolate it into something very meaningful. And I think what they do with this data is going to be -- I predict that it will be extraordinary. It will be helpful, it will be illustrative and it will change the way we think of imaging.

I know -- if you were on X for 5 days after Midjourney's announcement, this was a top 5 trending topic for 5 days in global communications to actually talk about this because there was a huge debate between technologists and the medical community. Medical community thought and think this is a bad idea because you're going to run into your doctor with every scan and you're going to take up their time with a bunch of unnecessary stuff. I think it's going to be the exact opposite. I think it's going to inform patients. I think more data -- and thinking that people are dumb is not a good platform to stand on. Being more educated, being more aware, having more data and then having intelligence that can extrapolate data to help you make the best decisions I think, is the future of health care.

And I'm just in awe of Midjourney's boldness and their vision, and they pretty quickly networked 40 of our devices into a single image. And normal MRI is about 40 teraflops of data, if that's -- if I have that correct. One Midjourney scan will be 400 teraflops of data for every one of these low-cost, inexpensive scans. So I think it's one of the steps that are going to make a massive impact on humanity.

Yes. And I mean, for you all, beyond the economics, I have to imagine the strategic value of the partnership and demonstrating in particular, what ultrasound on a chip can enable was probably one of the most interesting things as you've seen the excitement around the announcement and also your relationship evolve over time.

So John and I have been working really hard to communicate to investors what we think are going to happen in the future. We do the best we can to indicate what's to come. And we do that very mindfully on a quarterly basis and at times like this. But we do it in a manner that when we say we're going to do something, we're going to deliver on it. And if you go back in time, you'll see over the last several years, we've delivered everything that we've told you that we're going to deliver.

And seeing Midjourney come out in the type of way they pronounced the announcement, many of the people who've been following Butterfly have come back to us and said, "Oh, now, I understand." Now -- and we try to do as much as we can with direction and breadcrumbs. But yes, I think when that came out, it opened the minds of investors like, "This is real, holy cow." But it also opened minds of many other technologists in all different walks of life.

Merge is a Sam Altman company. Neuralink is a $9 billion start-up from Elon Musk to do brain-computer interface. Sam Altman is now starting his own Neuralink and he chose Butterfly. So I mean, this is -- we're dealing with some very, very smart people. And I think -- and David Holz at Midjourney is a complete genius. And for us to even be in a ZIP code of this brain power, it means we belong, first of all, but we're going to learn and grow, and there's a lot of other things that are going to follow.

Yes. And you've touched on Merge a bit. I do want to go into some more details around that because it takes Butterfly into a completely different frontier with BCI. What do you think about the opportunity itself? And what does it do in terms of validating the fundamental capabilities and potential of the Butterfly platform?

I think their selection of us is enough. I don't -- what I can't really do is go into what they intend to do other than what they've stated publicly. It's very, very sensitive stuff. But as I said in the beginning, very heavily negotiated. And if they didn't need us, we wouldn't be there.

And so it's a validation of what we've accomplished. And it's a validation that they believe where they want to go will be a lot faster if it's with us. And we are so honored. There's 7 founders in the business, each genius, the next to the next genius. And it's been -- it's just an extraordinary for us to, again, rub elbows with a lot of these future leaders and to be considered a core part of this technology.

The concept of -- there's going to be -- we just held a BCI conference in the Bay Area a couple of weeks ago. And what I learned, a lot of what people think about in brain-computer interface has to do with restoration of function, especially from an accident or chronic disease. You lose function of your arms. So now you can direct the robot to feed you or whatever, you can move a cursor on the screen. And I think that's where a lot of the initial BCI attention was. I think a lot of the future attention is going to be how do I improve my function as a healthy person. And there's going to be a lot of cool things and a lot of ideas as to living more productive lives and whatnot. And when you look at the consumer TAM of some of these ideas, it's kind of hard to calculate. It's so big.

So it's a major validation of our platform. It's a major validation of our IP, and it being now the fourth BCI company, and there's others we're still talking to, that will be a major category for Butterfly in the future. But I just think that's one category of a universe of other possibilities.

Yes. And I have to imagine being in a partnership with Merge also creates technology advantages for you all, meaning, is there lessons learned around imaging, sensing, software or AI that's also helping you continue to evolve the technology?

Yes. Yes. And that's why we've taken the core technology into the center of the business. And we've put the best minds in the center of the business. And whatever point-of-care ultrasound needs, we give it. But some of our partners are asking us to do things on the edges of what we do today. And they're pushing us in this realm, whether it's in longevity and battery life or frequency or voltage, they want to do all these different things, and it's testing us and it's pushing our boundaries into places.

And at times, as we've thought through something new, a new partner comes in with a very similar requirement. And now that other partner helped us kind of push ourselves to a place. So yes, I think -- and the beautiful part is the way we structure our partnerships is if there is unique technology that's developed in a certain area, in a certain field for that partner, that partner owns it. But for the outside that field of the area, it comes back to our core. So our brain is just getting smarter and smarter based upon being tested by all these extraordinary use cases.

And so far, all of the core tech that we're developing is fungible. So we're not developing a unique chip for a partner. It's all how do we make the next technology better and then it benefits everybody. There may be a day where someone says, "We'll pay you this to go here, and we want to just for ourselves," and that will have to be an extraordinary -- an enormous deal, but we're fully willing and capable of doing it. But today, so far, everything has been leverageable and fungible.

Okay. Wonderful. I'm going to switch topics again to go to P&L and John, maybe you spend a little bit of time around...

All the fun stuff.

Exactly. I mean a 20% growth trajectory is very impressive. I think investors are always keen to understand what is durable recurring revenue from partnership versus -- and transaction-driven revenue?

Yes. As I mentioned, the beauty of the partnerships are -- these are longer-term partnerships. So they are enduring and recurring. I mean what we did with Midjourney is a 5-year agreement, $74 million, and that's really the floor of what we expect to get from that agreement, both in terms of years, we do expect that that's going to be extended as well as in terms of what we can get from development, from chip sales, from revenue share as well.

On the other side of the business, on the core business, really looking at our pipeline. As Joe mentioned, we're moving upmarket to -- we're still selling across our e-com platform, but we're also moving upmarket with our Compass AI software. And so from that perspective, you pay really close attention to the pipeline and how that's accelerating as well as how we're landing and expanding. So revenue per customer is pretty key in how that's growing as well.

Okay. Wonderful. I guess maybe to wrap, what do you want investors to understand about the model today? And how do you want that to evolve over the next few years? If you can just give us a couple of KPIs that would be...

Yes. We touched on it briefly, but it's really the operating leverage that we have and the portfolio of opportunities that we have. I mean on the core side of our business, we're not just a one-trick pony in terms of probe sales. As I mentioned, we have Compass, we have the launch of our home business in October. We're coming up with 2 new products -- out with 2 new products next year. Early in the year, we'll have our next-gen probe and later in the year, we'll have the station.

And similarly, on the Embedded side, we're not a one-trick pony. Yes, the Midjourney deal, I think, certainly put us on the map. Maybe people may not remember what we can call it our Sally Field moment where we're somebody now. Somebody remembered. So there, in addition, with the Merge partnership, there'll be more partnerships to come, looking at that pipeline and how that's developing as well. So tremendous operating leverage, and we have significant portfolio opportunities that are coming off of a core platform of investment.

Wonderful. Well, we're out of time unfortunately. I could spend all afternoon digging into your brains around the opportunity set here. But Joe, John, thank you so much for taking the time to speak with us and giving us that detail.

Thank you.

Thank you.

Appreciate being here.

Good morning, or good afternoon all, and welcome to the Butterfly Networks Second Quarter 2026 Earnings Call. My name is Adam, and I'll be your operator for today. [Operator Instructions]

And I will now hand the floor to John Doherty to begin. So John, please go ahead when you are ready.

Good morning, and thanks to all of you for joining our call today. Earlier, Butterfly released financial results for the second quarter ended June 30, 2026. We also provided a business update. The release, which includes a reconciliation of management's use of non-GAAP financial measures compared to the most applicable GAAP measures, is currently available on the Investors section of the company's website at ir.butterflynetwork.com.

I, John Doherty, Chief Financial Officer of Butterfly; along with Joseph DeVivo, Butterfly's Chairman and Chief Executive Officer, will host the call this morning.

During today's call, we will be making certain forward-looking statements. These statements may include, among other things, expectations with respect to financial results, future performance, development and commercialization of products and services, potential regulatory approvals, revenue attributable to Embedded partnerships through revenue share, chip purchases or otherwise and the size and potential growth of current or future markets for our products and services.

These forward-looking statements are based on current information, assumptions and expectations that are subject to change and involve a number of known and unknown risks, uncertainties and other factors that may cause actual results to differ materially from those contained in the forward-looking statements. These and other risks are described in our filings made with the Securities and Exchange Commission. You are cautioned not to place undue reliance on these forward-looking statements, and the company disclaims any obligation to update such statements.

As a reminder, this call is being webcast live and recorded. To access the webcast, please visit the Events section of our Investor website. A replay of the event will also be available on this page following the call.

I would now like to turn the call over to Joe.

Thanks, John. So good morning, everyone, and thank you for joining our second quarter 2026 earnings call. What a great quarter it was. We closed above the top end of guidance, delivered record revenues, gross margins and adjusted EBITDA, all above where we thought we'd end up. As a result, we are raising our full year guidance, which John will share in his section.

Last quarter, I shared that three engines of growth were emerging and beginning to reinforce each other: POCUS; Home and Community Care; and Embedded. What's unfolding across our business this quarter shows that, that flywheel is now starting to turn. And I want to walk you through each one, starting with the one that's been capturing global attention, Butterfly Embedded.

The world now understands why we've been so eager to share details about the Embedded partnerships. We can finally unpack what a couple of our partners have discovered is possible with our technology. Our vision for a new category of chip-based ultrasound is unfolding right before our eyes. Midjourney's groundbreaking announcement to build a whole body scanner shocked the technology world and challenged conventional thinking and medicine. The promise of semiconductor-based imaging had never truly been appreciated until now.

As I've said to you many times, we're entering into an era where ultrasound will be liberated from big imaging and expanding beyond the four walls of the health system and moving to where people are when they need it. Butterfly's Embedded actually amplifies Butterfly's core mission of democratizing imaging by partnering with companies who share that same vision. Whether it's a handheld probe, a wearable device, or a whole body imaging system, one constant remains true.

Safe, low cost imaging paired with massive data capture in the AI era will transform healthcare. This is a chance to empower patients to better manage their wellbeing while creating one of the most important advances in improving human health. So no patient should be surprised with their health.

But six days ago, I received a call from a dear, dear friend of mine telling me that he was just diagnosed with Stage 4 colon cancer, and it had metastasized to his lungs and liver. I mean, just diagnosed? Stage 4? Are you kidding me? Do you know how long it takes for that primary tumor in this colon to develop and metastasize? Could be years. Where was his MRI? One of our employees was diagnosed with Stage 2 pancreatic cancer just by chance. Had she not had a scan for another ailment, it never would've been found, since pancreatic cancer is also asymptomatic and normally found when it's too late. But now with an unusually early diagnosis, she has a chance.

These stories, unfortunately, aren't rare. They're happening everywhere, every day to people we know and people we've never met. The recent untimely passing of Senator Lindsey Graham from an undiagnosed aortic aneurysm, something that could be treated with life-saving surgery if caught in time, is a sobering reminder that even people with access to the best resources in the world can fall through the cracks of the system. It's not about the imaging modality. MRI is the best imaging in the world, no question. But it's about the system, the cost, the barriers, the stakeholders, and frankly, the old thinking that screening is too troublesome for the current system.

Technology is changing that. Semiconductor-based imaging, mass data processing, storage, and AI will change that paradigm. It'll give you greater awareness of your health journey, help you make better decisions with your life, and be the scribe to your personal story.

Another key learning this quarter is how well-suited Ultrasound-on-Chip is for neurotechnology. Probably one of the most profound uses of our chip will be the ability to image, to map, to understand, and potentially apply therapy to the brain in the future while pushing and pulling information from it.

Aleph Neuro showed us two major breakthroughs in the last 2 months. First, they demonstrated a remarkably high-resolution 3D transcranial ultrasound of the brain using a contrast agent, showing not only that imaging through the skull is feasible, but that it may ultimately offer diagnostic insights approaching those of functional MRI. Next, they revealed a new way to use ultrasound to track the movements of the tongue to discern speech. Each manifestation of our technology can have global impacts to health and to quality of life. And this is just the beginning, everyone.

Before I move to POCUS updates, I want to mention that we've signed 2 more Embedded partnerships in the second quarter, bringing the total to 11. Two very exciting additions. One, another BCI company developing a novel application, and another pursuing a wearable for one of the most common areas of women's health.

Shifting to our POCUS business, we had several key accomplishments. We signed six more enterprise software deals deploying our Compass AI software to enable system-wide POCUS management and adherence. Many of these deals were paired with probe sales and we expect as the software is implemented, the growing base of probes will steadily follow into these systems.

Since last year, our Compass AI software pipeline has increased five times and we are seeing the benefits of that now. Compass AI is how POCUS becomes part of the healthcare workflow, driving adoption, utilization, and earlier diagnosis. I'm very happy with the team's progress. And I'm equally excited about our opportunity in government sales, including VA, DoD, HHS, and other federal agencies.

In the second quarter, we received provisional authorization to sell to the full VA, and we anticipate full FedRAMP certification in the third quarter, which will make us one of only 500 companies certified. The largest purchaser of U.S. healthcare is the U.S. government. And this channel will soon be open to us. Our government pipeline has grown in anticipation of this accomplishment and we are ready to execute on the significant opportunity ahead.

Medical schools also continue to be an important area of strategic growth for Butterfly. Today, we have a presence in more than 80% of the U.S. medical schools, and we are increasingly seeing institutions move from purchasing a handful of devices to implementing a one-to-one ultrasound curriculum.

Schools like Kansas City University are leading the way with first-year students now completing their entire ultrasound curriculum on Butterfly ScanLab platform. As that inaugural class completes its first year, we're already seeing insights that are helping other institutions build similar programs.

In Q1, we told you we placed nearly 1,000 probes in the six medical schools. In Q2, we more than doubled that through four additional institutional partnerships. As these programs have expanded, we've started asking a bigger question. So what happens after graduation? A recent survey we conducted across our medical school accounts found that while more than half say graduates frequently continue using POCUS in practice, nearly 3/4 have no pathway for graduates to keep a device into residency, and nearly 60% don't track their graduates at all. That reinforced our belief that there's an opportunity to better understand and better support new physicians as they transition from POCUS training into clinical practice.

That brings me to some exciting news this quarter. We entered a new partnership with the Edward Via College of Osteopathic Medicine or VCOM. It's a novel partnership model in which not only will every VCOM student have a probe to train with, but we will partner on a long-term longitudinal insights program.

As all of you know, Butterfly devices are connected to a cloud which stores image data much like your iPhone does with iCloud. We gather performance metrics on all users that help us understand usage patterns, locations, and the types of scans being performed. We're not quite ready to share the detailed logistics of the program yet, but we have an exciting plan to track and learn from the ultrasound journey that follows a one-to-one medical school program.

VCOM made the investment to provide every incoming student with iQ3 devices, while Butterfly subsidized the initial distribution of iQ+ devices to get the program going. You'll see our investment reflected in the lower ASPs we've reported this quarter. We believe that's exactly the right investment to make as we establish what we think is a groundbreaking new model for medical education. This partnership represents the potential for more than $10 million in total contract value over the next 4 years.

So regarding international, you'll see that our team had a down quarter. We believe it's really no big deal as our distribution sell-through is up meaningfully year-over-year. We had a handful of larger orders shift into quarter 3, so we expect a much stronger third quarter ahead. So John will say more about this when he covers the financial results.

One highlight we're particularly excited about is Brazil, where we recently achieved regulatory authorization. Brazil is the largest medical device market in Latin America and one of the fastest-growing ultrasound markets in the world. Meeting Brazil's regulatory requirements is a great validation of our technology and opens the door to a significant long-term growth opportunity for Butterfly in the region.

So I'll close out POCUS with Butterfly Garden. As planned, we've officially released our proprietary beam steering capabilities, including iQ Slice and Fan, to both our Garden and Embedded partners, furthering their development opportunities.

In the second quarter, we added a new Garden partner focused on precision needle-guidance with AI, and we expect three of our existing partner tools to become commercially available by year-end. Importantly, those three partners will launch with direct access into Butterfly's cloud, allowing their AI results to flow right into our users' existing clinical workflows without toggling to a separate application. It's an exciting milestone as Garden evolves from a developer ecosystem into a commercial AI platform.

So before I turn it to John, I want to comment on our Home and Community Care business. We are ready to begin managing our first commercial state on October 1st. Over the third quarter, we'll be training all the participating nurses in the state while we finish setting up the necessary infrastructure. We are excited for this to transition from a pilot to finally implementation. We will do everything in our power to support the success of this commercial execution and plan to discuss extending this offering to other states in the first half of 2027. We're also now in conversations to expand this service to other key partners. So Butterfly Home and Community Care is finally here, and we expect the revenue in the fourth quarter of 2026.

With that, I'd like to turn it over to John to discuss the financial details. John?

Thanks, Joe. Butterfly continued its strong and focused execution in the second quarter of 2026 with an increase in revenue driven by strong growth in our Embedded business.

Core POCUS is also solid, with performance reflecting our increased and expanding focus on the enterprise market. Gross margin increased, and we saw continued improvement in operating performance with further reallocation of resources towards higher ROI opportunities and markets.

Building on the operational highlights Joe just outlined, including the large TCV contracts for Compass AI and medical education and progress towards commercializing Home and Community Care, let me translate these into the financial highlights in the quarter.

Record revenue attainment above the high end of consensus and above the top end of our guidance. The highest gross margin for Butterfly in its history. Adjusted EBITDA, that was above consensus and our guidance range, with improvement driven by our revenue performance, higher gross margin, and continued financial discipline. And growth in probe unit sales of 23%.

With that, let me move on to our results. Second quarter revenue was a record $32.6 million, an increase of 39% year-over-year. Our growth was primarily driven by Butterfly Embedded. In addition, in core, we had strong performance in U.S. health systems and medical schools, including a large sale to VCOM, which helped to drive strong year-over-year unit volume growth, along with Compass AI sales gaining traction.

Breaking things down between the U.S. and international channels, during the second quarter, U.S. revenue was $27.6 million, which was 57% higher year-over-year, driven by revenue from Embedded as well as solid demand in the core business with unit sales up 23%.

As Joe mentioned, international was down year-over-year. Total international revenue decreased by 14% year-over-year to $5 million in the quarter. We faced a tough Q2 comp against last year's record international second quarter, aided by tailwinds from the iQ3 launch. We also had a few deals expected to close in the second quarter move out into the third quarter. We expect improving performance for the balance of the year and we are excited about our recent launch into the Brazil market and continuing to expand our global footprint.

As I did last quarter and will continue to do going forward, I want to provide you with the revenue split between our core POCUS business and Butterfly Embedded. We have also included this split in our 10-Q. As a reminder, the core business includes probe sales and related software, Compass AI, other services, and in the future, Home.

Embedded revenue currently includes one-time NRE payments, annual license fees, revenue from SOW-driven development work, and chip sales to Embedded partners. With that, revenue for the second quarter was $21.9 million, an increase of 2% versus the second quarter of 2025. This increase was driven primarily by growth in volume in U.S. health and medical schools.

Of note, and as Joe mentioned earlier, the core result includes the execution of a new strategic agreement with VCOM. This agreement is for 4 years with a potential TCV exceeding $10 million. This also included preferred pricing for the iQ+ to help secure the larger multi-year commitment to the iQ3. This investment in the alliance with VCOM is a template for Butterfly going forward to help deplete our existing iQ+ inventory, while enabling broader and deeper penetration of the medical school market ahead of the launch of our next-generation probe in 2027.

Sales of the iQ3 and iQ+ in the quarter were up 17% and 33% respectively year-over-year. The higher year-over-year mix of the iQ+ in the quarter and the VCOM deal result in a 12% decrease in the ASP. Butterfly Embedded revenue was $10.8 million, an increase of 439% versus the second quarter of 2025. This increase was primarily driven by the Midjourney partnership.

Moving on to gross profit. Gross profit was also a record at $23.3 million in the second quarter of 2026, a 56% increase as compared to the prior year gross profit of $14.9 million. Gross profit margin percentage increased to 71% from 64% in the prior year period, a 7.7 point increase. Gross margin percentage was positively impacted by the increased mix and higher margin of Butterfly Embedded revenue.

Moving to EBITDA and cash. For the second quarter of 2026, adjusted EBITDA loss was $1.4 million, compared with a loss of $6.2 million for the same period in 2025, an improvement of 78%. The improvement in adjusted EBITDA loss in the second quarter was driven by contribution from higher margin revenue and continued financial discipline.

Our cash and cash equivalent balance, excluding restricted cash at the end of the second quarter, was $125 million, and the use of cash in the quarter was $13.3 million. This compares to use of cash of $7 million in the prior year quarter.

We are well-positioned, as we move forward, to continue to invest in our business in areas where we see significant opportunities for additional growth and disruption, which includes: scaling our POCUS business through global market and market segment expansion; continued penetration of Compass AI as a core operating system for health systems; app releases in Butterfly Garden; and 2027 new product launches; continued AI and semiconductor innovation with the development of our fourth-generation chip; broadening Butterfly Embedded partnerships; and moving Home and Community Care to the commercial phase.

Before turning to guidance, I want to update you on the general global macroeconomic environment relative to Butterfly. We continue to monitor the conflict in the Middle East and the ripple effects on the global economy, as well as pressure from tariffs in certain markets and AI-driven supply chain shortages.

We have mentioned in the past that there are some impacts to our business. However, they have been minor and we continue to manage through it and make the appropriate adjustments. Our second quarter 2026 results are indicative of this, and our third quarter and full year 2026 guidance include any expected impacts.

I would now like to turn to our outlook for the third quarter of 2026 and for the calendar year ending December 31, 2026. In the third quarter, we expect revenue in the range of $26 million to $30 million for a year-over-year increase of 30% at the midpoint. We expect an adjusted EBITDA loss in the range of $6 million to $9 million. For the full year 2026, we are increasing our guidance for both revenue and adjusted EBITDA.

We expect revenue to be between $119 million and $123 million, an increase of approximately 22% to 26% over 2025. We expect our adjusted EBITDA loss to be between $19 million and $23 million. Our guidance for adjusted EBITDA in the third quarter and full year includes increased investment in key areas to support continued innovation and revenue growth in our core business and our emerging Embedded business for 2026 and beyond.

In summary, we had a record quarter. We had our highest revenue for a quarter ever, and we beat both our revenue and adjusted EBITDA guidance. We are very well-positioned as we head into the second half of 2026, and as the raise to our full year guidance reflects, we look forward to continued growth this year and beyond.

We remain focused on gaining share through deeper penetration of existing customers and adding new customers in our core POCUS business and expanding our Ultrasound-on-Chip partnerships. The business continues to get stronger across core POCUS and Embedded. This is happening while we continue our intense focus on driving operating efficiency across the business and return on investment. I continue to be excited about what is ahead for the company in 2026 and beyond.

Now, let me hand it back to Joe for some closing comments.

Thanks, John. Looking ahead, we are making progress on all fronts in our mission to democratize medical imaging. The Midjourney announcement was captivating. Millions of people have viewed the announcement, and it was a trending topic on X for more than a week.

Technology companies of all sizes have taken notice, and while we've always believed Butterfly was key to these conversations, we now have a spotlight in the discussions around AI, compute, and the many applications of Ultrasound-on-Chip technology. It's been an exciting last 60 days, and we finally can communicate with investors and partners about the future our technology can open.

What's especially exciting is that our semiconductor roadmap is advancing not only in image quality, but also the ability to generate large data sets with our high-speed Apollo chip and push data directly into GPUs for large-scale AI processing.

As new compute architectures emerge, our partners envision converging image acquisition with AI-driven data interpretation, and this vision is becoming increasingly achievable. Our Apollo platform is moving forward on schedule. We've been investing in our chip to meet the demands of both our Embedded partners and the POCUS business. We've now completed our primary design and are transitioning beyond research and into the development phase.

We're on track with our scheduled commitments to deliver Apollo chips to Midjourney on time, by the end of 2027, beginning of 2028. Just as exciting, we're continuing to make excellent progress on our next generation probe that'll be built on our P5.1 chip. The first full prototype images are outstanding, and we're on track for our first quarter 2027 launch.

iQ3 already delivers best-in-class image quality and will remain our workhorse, while the P5.1 builds on that foundation with the first CMUT chip to deliver harmonic imaging. This allows an even higher level of performance that's important to certain specialties like cardiology. What we've seen so far gives us tremendous confidence in what's ahead.

Our P5.1 chip is showing us better images, specifically in cardiology, than we see from our PZT handheld competitors. If this holds true, with better imaging in a handheld than our competitors, both big and small, why would anyone buy anything other than a CMUT? We are preparing for a big launch in 2027, and we'll continue that momentum by entering the cart business with iQ Station later in the same year.

The flywheel is turning. We've increased our guidance for 2026. We're adding Embedded partners more frequently, and we'll be launching some key Garden partners' AI tools in the second half of the year. By early 2027, we are planning for our sales force to have what's expected to be the best handheld imaging in the world, Compass AI, Garden AI apps, and then we'll march upstream into the health systems with the most holistic ,integrated POCUS platform to begin tapping into the existing multi-billion dollar global ultrasound cart market in health systems.

On top of that, the Home and Community Care will be going commercial. It's all coming together. I cannot be any more excited about our future.

So with that, Operator, please open it up for questions.

[Operator Instructions] Our first question today comes from Josh Jennings from TD Cowen.

Congratulations on a strong quarter. Continue to add layers of revenue opportunities onto the story. That flywheel seems to be turning, as you called out. Wanted to touch on Embedded and then also just the POCUS core growth. Just on Embedded and just revenue visibility with the Midjourney and Aleph announcements, how should we be thinking about the cadence of Embedded revenue over the next 12 to 24 months, and should we expect meaningful contribution from additional partners beyond Midjourney before the larger chip and revenue-sharing opportunities materialize with that agreement?

Yes, Josh, I'll take the beginning of that one. As we've said in the past, we expected Embedded to be somewhere around $7 million to $8 million per quarter, and that would fluctuate a bit up or down $1 million or $2 million here or there. It was down a bit in the first quarter from that. Obviously, it was up in the second quarter. And we expect to kind of maintain that cadence. Obviously, as we add additional partners and we go broader across Embedded, I would expect that to increase, but that's where we are right now.

Yes. I mean, all I'd add is we added two additional partners, and each of these are -- have a lot of upside to them. And so they're all research programs. They're programs where people are looking at building new businesses, and at any point in time, they can reach a milestone and want to go commercial, and that's a second bite of the apple for us. So our revenue in Embedded includes all of our partners, not just Midjourney, and so when we add people, there are revenue in the numbers. But it's when they go commercial that we can get the big pops. And there'll be some stuff coming up in the future. We'll be adding more partners, and I'd anticipate some of our current partners going commercial. So this is just going to layer on top of itself. It's going to grow and the Midjourney announcement has also opened a lot of new conversations that are very, very exciting.

Yes, I wanted to touch on that. Just the level of inbound interest accelerating following the Midjourney public demonstration, the Aleph announcement, 11 Embedded partners. How many additional partnerships can the organization realistically support simultaneously? Embedded is -- if maybe just help us understand, can you get out to 20 partnerships plus over time and support all of those efforts?

Yes. I think -- I don't see in my mind at the moment a limit to the partnerships. Because I think -- the beautiful part about this is that there is commonality. There is a flywheel. So we're not developing new chips for each person and then taking on all this cost and supply chain and CapEx. It's the same chip. What we're doing is building a software platform that allows them to do the work for themselves. And so -- and then sometimes they'll ask us to do specific work that's right on the fringe, and then we'll do it for them.

But I see this evolving into something that's pretty standardized. People come in, they license our technology, they purchase our hardware. And then at some point in time, they buy chips and get licenses for software. And then if they wish some market protection and wish to purchase exclusivities, then that's when those deals get bigger. But when I look at companies that we emulate, they have standard pricing and standard business operating procedure. And it's all into our current semiconductor platform. So I don't see a limit to this. I don't -- We're not a job shop. I see this as a scalable, core part of the future of imaging technology.

Yes. And one thing there, all the development work we're also getting compensated for, so it's not as if we're doing that for free.

Yes. And I think -- Thank you for saying that, because I think what we've done now is up until, I think the end of 2025, we were multitasking. We would go into our core development teams and have them do things for partners. What we've done since the end of 2025 is kind of firewall off and build an organization that are dealing with our partners, and we're going to continue to do that, and it'll scale based upon our partner revenue. So it doesn't affect our core business. We're continuing to make sure that Embedded grows at Embedded's pace and POCUS grows at its own pace, and Home will grow at its own pace. So we're becoming more mature and sophisticated, and we're focusing on scaling ourselves to -- I think we're going to see years and years of growth. This is a highly, highly scalable opportunity.

Outstanding, and sorry to ask one last question. I was talking about, I think POCUS core growth in my initial inquiry, but I wanted to actually touch on the Home transition and moving through commercialization. I know you've talked in the past about that being a big revenue opportunity. I'm not sure if you reiterated anything, if I missed it on the call earlier, but, how should we be thinking about revenue contributions once you launch in 4Q? Is that baked into guidance -- the updated guidance that you put on the tape today? And just how can that build out, and then maybe what is the TAM there? I think it's pretty large.

So, yes, Home is built into our guidance, and obviously, it's relatively small launching and expected to launch in October and it will build from there. We have multiple millions in 2027, but for this year, it'll be certainly short of that since it's only 2 months.

But it is a very large opportunity, and it's a function of how successful we are. If the success in our pilot continues in our first commercial implementation, why would anyone not bring it everywhere else in the country? Because we're helping patients stay in these skilled nursing facilities, reduce the revolving door into health systems, reduce the overall cost of care of these patients. And I think there's a whole flywheel on its own sense as it becomes understood how impactful it is to empower nurses with AI at the bedside to keep patients properly diagnosed and properly medicated as to where they are in their condition. I think it'll not just be congestive heart failure, it'll be a lot of other modalities.

And again, this whole thing ties together because as new Garden apps come out, like right now with HeartFocus, they not only can do a B-line scan, but they'll be able to do a cardiac echo at the bedside. And then in the future, they'll be able to do other types of things like deep vein thrombosis or bladder or others. And so it's not just about expanding it to more and more at-risk providers. It's also about expanding the use case within those providers and building this out as a whole modality that allows people to help patients where they are.

And again, that's a part of our overall mission. So this is a flywheel. This does compound on top of itself. And the center of that bullseye is our Ultrasound-on-Chip technology. And then as we build out new AI capabilities, that grows the user base, and as we grow the user base and we have all these new outcomes, then it grows the customers to do more and more things. So again, this is just the beginning, and it's -- all the things that we've been working on, we're kind of executing across the board, and it's going to build into something meaningful. So if -- we don't like to commit on deals that we don't have signed, but the opportunity in 2027, I think, is real. But we have to execute now. We have to delight everyone. We have to show that this works and that it's scalable for them. And if that happens, it gets real big, real quick.

The next question comes from Chase Knickerbocker from Craig-Hallum.

Just a few questions from us. On the VCOM partnership, can you speak to how the partnership with VCOM works in terms of revenue recognition, how that flows into the Embedded business? And going forward, is there an opportunity to sign more of this type of deal?

Yes, let me just take the very first part of that. So the VCOM deal was purely core. It had nothing to do with Embedded. It was really the sale of iQ+, iQ3 and related software over an extended period of time.

Did that answer the question, Chase?

Yes.

Yes, I can talk more about the...

[indiscernible].

Yes, absolutely. So VCOM has a desire for all their students to be able to learn ultrasound and have their own probe. I think there's a growing general consensus that when you have your own probe, you can scan yourself, scan your friends. You have opportunities to really acclimate and become educated. And so, VCOM has made a long-term commitment on an annual basis to now stock the first-year students with the state-of-the-art technology and brand-new probes.

And in order to help with their objectives of having more of the upper classmen have access, we've used our iQ+ platform, and we've subsidized those sales into those upper classmen. So we make it easier for them to onboard into this model. They didn't just want to have a part of the school in it, they wanted to have the whole school in it. And so, we've made that available. And then what happens on an annual basis is, they keep on staffing or supplying the next or the first class with the most recent technology. So it's a great, great, great signal to the overall education market. It's a great validation of our mission, our joint mission.

And there'll be another announcement in the future. They have a very novel idea that we've agreed with fundamentally and that we will be supporting, but we're going to allow them to announce it to their constituents and to their students. And then when they announce it, we'll educate you all on what they're doing. But because of our subsidy of the iQ+ probes, that's what kind of was reflected in the accounting of our ASP, which is -- again, for us, it's the best way to make an investment, which is to put more devices into more students' hands.

Thank you for that additional detail. Just one more follow-up. Going back to Embedded. Could you give us some color on the components of the rev rec in the second quarter? I just think that would be helpful for modeling going forward.

Yes. Just -- We've kind of laid this out before. There's multiple parts of this. There is the initial upfront payment, which was 4 -- $15 million. There was $10 million per annum for license fees, and then there was the development work with chip sales and revenue share to follow once they commercialize. We recognize that revenue as we do work against certain milestones. So as we progress against work that's related to the project, we recognize the different pieces of that.

So in the fourth quarter of last year, when we initially signed the deal, we didn't recognize all the $15 million upfront. Obviously, we got the cash, but we released that as we do work against the overall project, the overall program. We made a good amount of progress against some of the milestones in the second quarter, so that resulted in the Embedded number being higher than, say, it was in the first quarter.

The other thing I want to highlight on this, so the overall contract that we announced back in November of last year for $74 million, I wouldn't say, that's not it in total. We talked about some of this in mid-June after Midjourney had their event out in California. There is opportunity for us to certainly make additional revenue on top of that $74 million over time, but subject to the amount of work we're doing, if we do additional work related to our next generation chip, and of course, when they commercialize the chip sales and revenue share. But for now, it's effectively us doing work against requirements for working towards commercialization with them. And that allows us, as we look at that and measure the work that we're doing, to release revenue in the respective quarters that we do it in.

And then just to add on that, we have 11 partners. That's one. All the other 11 are -- there's revenue on software licenses, there's revenue on hardware purchases, there's revenue on semiconductor purchases, and that grows as we grow our partners. And also, as our partners are in different phases, they will then buy more and do different things. Also, we have -- of our other 10 partners, we do work for them also, and that revenue rolls into the number. So -- and that grows based upon where they are in their development cycle and also as we add new partners.

The next question comes from Ben Haynor at Lake Street Capital Markets.

First off for me, on the training and Community Care commercial launch here coming up, how large is the group that you need to train? How many patients does this plan to touch in the initial commercial state?

So in the first state, I think we're training between 15 and 30 nurses. I don't know the exact number, but it's something within that scope and size. And right now, until the program gets bigger, I don't think we're going to disclose the number of patients under management. I think we want to get a little more mature and a little bit more consistent before we do that. And when revenue becomes meaningful, we'll translate where our revenue model is and what the scope is. But right now, I think the first state is between -- these are nurses who work in skilled nursing facilities full time, and I think it's between 15 and 30, I think is the right number.

So it doesn't sound like a situation where you need to add sort of immense training capacity, and this is something that could go nationwide or much broader without too much on the cost side for training purposes?

Yes, well, I mean, training is a core competency of us and we're very fortunate that one of our best internal ultrasonographers has taken on the role of going full time into Home and Embedded, and she's going to do a great job. And as we scale, we have actually one of the -- we have a great partner. We have several great partners in education and medical education. We leverage our internal team for as much as we can do. But if we also had the wonderful opportunity to go into many states quickly -- we have partnerships with state-based education organizations that would be -- And this is only one use case and one AI tool. So the bar is much lower than what a normal training would be.

So the ability for us to scale quickly based upon all the nationwide partnerships we have with individual training organizations, and some of those organizations are national and global. We're very confident that we could do this deployment. And then once the deployment is done, we would make sure that we have the staff and the team in place to be able to maintain that, because training is not a one-time thing. This -- The value of this program is that we are with our partners every step of the way for their journey, for training new people, managing their data, managing data transfer, managing the clinical assessments and the whole thing. And so, we're eating our own dog food. We're using our own stuff to deliver value side-by-side with our partners.

Got it. And then maybe there's not an easy way to characterize this, but on the Butterfly Garden launches that are stated for later this year, is there anything you can share on kind of the commercialization efforts that you expect from these partners? I mean are they going to be making big splashes? Do they have large commercial organizations behind them? Any color there would be very helpful.

Well, I do -- I think each of these companies are smaller, growth-oriented companies. They're not these large companies as of yet. But we are going to help them become larger companies, we hope. And I think what we've been waiting for is their FDA approvals, and also they've made a lot of requests of us to integrate more into the platform. For example, while we've built an SDK that allows them to pull data for their AI models, we've had people say, "Well, we want to use your cloud to store the -- our own data, and we want to be able to do other types of scans and whatnot." And so we've been building out the robustness of our Garden platform in concert with where our partners wish it to be, and a lot of that now is kind of intersecting on the second half of the year.

And I think we're going to have three partners who will come online, and we are going to have -- I think that's going to become a cadence now because there's a lot of work happening. And I think it's -- there's -- what's not present in the marketplace with the consumers of the technology is how impactful AI will be for them. And so, I think each new app points to the entirety of the apps. And as you get two companies, three companies, five companies, 10 companies out there marketing their capabilities, it builds consumer awareness. It builds an identity that this exists, then there's more proof cases, and I think Garden creates its own flywheel. And I've mentioned in the past that I think this is the key to crossing the chasm, to really getting the user base of ultrasound, of point-of-care ultrasound to increase because the learning curve is so high for standard ultrasound.

But as there's more AI capabilities, people get more and more comfortable in using this every day. So I think that whole Garden flywheel as far as getting more partners in the marketplace, it turning into revenue, it turning into increased utilization, is now on the cusp of starting, and we'll have exciting launches now each quarter going forward into '27 and '28.

The next question comes from Steve Lichtman from William Blair.

Couple questions for me, Joe. On Embedded, you've been laying out in your investor decks the many potential application areas and use cases with your current and potential partners. As you look at those buckets, those different categories, are there some that you see as generally sort of earlier in the pipeline potential or sooner and others that might take more time in general, or is it pretty even across the board as you look out over the next few years?

Well, there are a few partners where it is a relatively low fruit. So our -- first of our two partners that we've announced publicly, one was Mendaera, which is a vascular robotics company. They were more interested in using our current sub-assembly of what we have in our tech to help create the type of image they need and using some capabilities that only exist within semiconductor-based ultrasound. And so they will actually be commercial soon with Butterfly. And over the next, I don't know, within the next six months, maybe sooner than that. I'll let them make their announcements, and I don't want to get too far ahead. But that is more about using our core technology to help amplify some really exciting novel things that they are bringing into the medical market.

And then we also have another partner, Sonic Incytes, that does fatty liver diagnosis. That is -- will be commercial, where the bar, again, is lower because they're using existing capabilities and implementing it into their systems. We then have a set of partners that are doing a lot of primary research. And that -- I would say, the -- I think it's four neurotechnology BCI companies now in the Garden, that they're doing work that is nuts. They're doing work where they are looking at ways of understanding the brain that have never been done before. And I think that has a longer timeline to seeing something in the clinical -- in the marketplace. Because there's not only the technology of feasibility, but then there'll be, I'm sure, a long regulatory path in order to get there.

Now, I will say a couple of people have more consumer-based ideas. Some are focused on chronic diseases, but some are focused on behavioral. And so, the bar might be a lot lower, but we'll see. And then some of them have really novel clinical applications that I think are probably easier to prove than the brain. They just have to be done. So I think there is for -- just like in any portfolio, we have some things that are pretty near, we have some things that are probably medium-term, and things are way out. And the stuff that's way out is, I think the opportunities are -- the markets are so large, it's hard to calculate. So it's a whole portfolio of things.

And then on top of that, we're talking to a lot of new companies in the last 60 days that we hadn't talked to before, that are very heavily resourced. And you never know if one of those catches on, it could be a Midjourney type of hit the ground running hard, and that can happen at any time. The Midjourney announcement definitely woke up or built awareness of our capabilities in the large healthcare community and also the big tech community. And while a lot of our partnerships so far have been with earlier stage companies, obviously Midjourney is a medium-stage company with a lot of revenue. We're talking to a lot of very well-funded companies and you never know, one of those can kind of change the calculus like Midjourney did.

And just secondly, on Home Care, I think you've talked in the past about the business model being different for Butterfly in that channel. How are you thinking about that revenue model and what it'll look like with these customers?

And then just secondly, what milestones are you looking for with this first customer to get you comfortable to kind of put the foot down on expanding to additional customers?

Two great questions. So the first question is simple. We will charge a fee for the size of the population we manage. We won't go into the mechanisms of that fee, but there will be -- as the patient population increases, the amount that we charge to manage that population will increase with it. So that's one set of consistent revenue. And then there will be a charge for every scan that's done, because there's a read associated to it and work that we have to do.

So there'll be kind of a fixed component based upon the size of the population, and then there'll be a variable component based upon how many of the scans that are done. And as that matures -- that model is just very simple. You add patients, then your fixed kind of revenue goes up. And then of course, as you're adding patients, your variable scan revenue goes up. And as we -- as that matures, we'll be able to kind of unpack that for you. What was the second half, the second question? First was the revenue. What was the second question?

The milestones. Just in terms of the milestones...

Yes. 100%. So within this patient population, we're focusing on congestive heart failure patients. Over the last decade, it's become really aware to fee-for-service payers that there's a revolving door between nursing homes and health institutions. And so, someone goes in a nursing home, they -- their care might not necessarily be consistent with their disease progression. They then have to be put in an ambulance, sent to the hospital and cared for, have their diuretic managed, and then they get sent back to the nursing home only three to six months later to have that occur again and again and again. And so there have been now capitated payments, there have been penalties, and there have been incentives on reducing the cost of care for those patients.

And so, what we are targeting is a reduction in admissions and readmissions for congestive heart failure patients. And nationwide, if you're a diagnosed congestive heart failure patient in a skilled nursing facility, you're going -- you're probably getting admitted or readmitted -- in 25% to 35%, you'll see a readmission rate or admission rate for that population. So what we are targeting is a reduction of that readmission rate. So we've set goals. We were very productive in the pilot, but it was a smaller, of course, population. What they want to see is, are we able to maintain the type of success of keeping patients healthier and in the bed and stable longer? And if those results translate -- I mean I forget the absolute number, but when you look at how much at-risk providers have to pay for readmissions, it's pretty astronomical.

I think our chronic care is, I don't know, 70%, 80% of our healthcare costs. It's -- the lion's share of healthcare is managing these chronic illnesses. And when you reduce a -- an admission or a readmission of these patients, it goes right to -- it goes -- I don't want to say it goes right to the bottom line because I don't fully understand, I don't want to profess their economics, but by reducing the readmission rate by 5%, by 10%, the economic impact to these at-risk providers is significant. And so, that's where we think -- as we sit back and we look at the macro picture of democratizing medical imaging and bringing medical imaging to the patient's bedside, we think that earlier diagnosis and managing patients where they are probably going to be one of the largest reductions in overall healthcare costs by simply democratizing earlier diagnosis.

And -- but the learning curve is high, the bar has been high, and it's a complete shift of behavior. That's why we are doing Home, because we've developed the skills and the capabilities of deploying this technology. So again, we're eating our own dog food. We're going to do it ourselves. We're going to take risks with our -- with these at-risk providers, and we're going to show them that we can reduce these readmissions. And when we do, we are -- there is a component where if we're successful in reducing it by a certain factor, then we are going to actually also be rewarded by a percentage of those savings. So -- And of course, if we don't, then we will -- it goes the other way, and we bear more cost of it too. So we are going with some risk, and we're doing it gladly because we understand what this technology is capable of doing.

So the moment these pilots and the early implementation actually starts showing the real cost savings, that's when these at-risk providers are going to jump in with both feet because that's just a lot of money that'll hit their bottom line by doing what we want to do and everyone wants, and that's everyone live healthier, better lives, be able to deal with things when they occur, empower caregivers where the patients are, and not have to have these significantly recurrent high-cost incidences. So this is a very big deal, because it's not just what it means in revenue to Butterfly, but I think we can reduce the market development time, and we can educate people on the power that this can have a lot sooner.

The next question comes from Raylin Konakuty, Freedom Brokers.

It was a great quarter, I can see. I have multiple questions as well. First question is about the next generation probe and the cart market entry. You have talked about the harmonics chip, its introduction at [indiscernible] and we are expecting the first launch in early 2027. And we can see that iQ Station is also targeting cart markets later that year. You said that once Butterfly achieves that harmonic imaging, there will be no reason to buy a single-dimension handheld device. This is a very, very great claim that essentially argues for market share capture across the entire central system. And my question is about the quantification of that. Can you please quantify the addressable market you are targeting with the new probe versus the current iQ3, and give us a sense of expected price point?

The -- Well, I have to tip the hat to my other analyst friends. That was the best question today. So -- fabulous question. A lot there to unpack, I'll do the best that I can. So first of all, when the digital camera reached five and seven megapixels and had an equivalent image to film, what happened to the film market? How -- Why would people still use an analog device from a handheld standpoint?

We've asked ourselves, once our image quality is better than other handheld PZT devices, like literally, why would you use them? If they're -- Whether they're three-in-one or they're specific to a certain application, they're not networked, they don't have the networked AI, they don't have 21 different presets. They have to have multiple -- they have to have multiple other things that have to occur for them. And they're also expensive. It's like, well, if our image quality is better, why would they ever use them? So I mean, we'll see what the market says, but in my extrapolation of where we're at, our next probe is just going to be better. And it'll be like, okay, I think we've reached to that point now, and that's why image quality is not going to be the factor after our next probe. It'll be AI processing and delivering more value at that bedside.

Now, today, the point-of-care ultrasound market in actual revenue is only $300 million, $350 million based upon a Signify report that just came out just a month ago. And we're the fastest-growing handheld company in the world based on that report. And so, how much more -- and so -- now we know the overall market for handhelds is, we think upwards of $20 billion if everyone gets a probe, but the existing market today is a $300 million market.

Now, the existing market for small compact carts and POCUS carts is about $1.6 billion to $2 billion. And that's how many -- what the cart existing market is. And that -- for a small company like ourselves to take the burden of market development, it's a lot of cost, a lot of effort, a lot of education, et cetera. But now to enter a market where we're just going to have a better product, it's an existing market. They -- The dollars are in the hospital's capital budgets. They want to do it. We're just going to walk in with a better product, and we're going to take market share. And that's what happens with iQ Station. So the first phase was making sure that we get our image quality to a certain phase. And then the second phase is we're going to have a better product offering than existing POCUS carts that are out there. Because those POCUS carts are kind of de-featured carts. They're not the type of network carts, and they're also certainly not networked on a one-to-one basis, too.

So our next product, I think, is going to generate a lot of revenue for us because we won't have the burden of developing a market. We'll have the benefit of coming in with a better product into an existing market.

I think the other upside revenue opportunity with the next phase of our imaging is we believe that our next probe is going to have a better cardiac image than the largest imaging companies have for their handhelds. And when people see it side-by-side, again, to have one probe that can do the whole body, that's networked, that has AI, that can merge into the health system's EMRs and all their data, and have a better image -- Before you'd have to have -- before you would have to give up something when you had something that did everything. Now, this probe is just going to have a better image, and so you can do everything and have the best image. Why buy anything else?

This was an answer that I didn't expect. Congrats.

All right, everyone. That's the last -- Operator, that was the last question?

Indeed.

Okay. Well, sorry, we're about 7 minutes over on the call. Just a lot. Thanks for all the questions for our analysts. Please excuse my enthusiasm, but the -- my first few years here were a pretty hard road getting things situated. But Butterfly has had a vision of democratizing imaging, of having every doctor, every nurse with their own probe, and that vision has increased, and our execution is right on. And so, I'm very excited about what we'll deliver into the future and just very much appreciate everyone's support. Thank you.

This concludes today's call. Thank you very much for your attendance. You may now disconnect your lines.

The contents of this call and the views expressed on it are solely for the use of the clients and embedded guests of TD Securities. Any publication, distribution, reproduction, posting, sharing or transmission of this information without the expressed permission of TD Securities is prohibited. As a reminder, we are not interested in receiving and you should not disclose any confidential information. In the event that you inadvertently disclose such information, please notify us as soon as possible. Okay. Go ahead, Josh.

Thanks, Carrie, and welcome, everyone, to another round of the TD Cowen Medical Devices Emerging Growth Call Series. We are excited to have the executives from Butterfly Network in our presence here on a big day after a huge night last night. Joe DeVivo, President, CEO and Chairman of the Board; John Doherty, Chief Financial Officer; and David Horsley, SVP of Innovation, Butterfly Embedded and Garden. Gentlemen, thank you for letting us host this discussion today in this event and for participating in our Emerging Growth Call Series.

Thank you for having us. Very excited to be here.

Absolutely. Well, Joe, you've been -- I guess not -- and warning isn't the right word, but highlighting this Butterfly Embedded franchise that was rebranded, I think, over the past year and some of the partnerships that were coming through, including Midjourney. And last night, there was a big unveiling event by your partner of the Midjourney scanner, and I think there's even more in that Midjourney medical pipeline. But maybe I can hand it over to you for just maybe some high-level foundation setting comments about the unveiling last night, and then we can dive into some more specific Q&A.

Well, Josh, first of all, we really appreciate TD Cowen's support. The bank has been -- your team has been amazingly supporting of Butterfly, and we appreciate, again, your personal support. I have great amount of affection and loyalty for people who were there early and were believing in us early and that was you. And so I just want to thank you for that and thank everyone on the team at TD Cowen for the support.

Yes, we've realized a couple of years ago that we are on to something special. We realized that our core Ultrasound-on-Chip technology that we developed for ourselves was very unique and can empower a lot of different use cases in health care and outside of health care. And we are very fortunate to have captured the imaginations of several very powerful, very creative and incredibly capable partners. And we've created 9 partnerships so far. We have a lot that we're working on.

But last night -- everyone did see in November an 8-K that announced a licensing deal, and it was a pretty substantial for a company our size, a licensing deal, which set out the framework of a partnership. But of course, that partnership, even though it was a material financial relationship wasn't ready to be communicated publicly. We are developing technology that allows semiconductor-based ultrasounds, software AI and all different types of components that not only help us in point-of-care ultrasound in our core business, but can be used by other parties to let their ideas come to life.

And our role as in Butterfly Embedded is to put our partners on our shoulders -- on the shoulders of our 600 patents of hundreds and hundreds of millions of dollars of investment in the semiconductor and all of our hundreds of millions and millions of clinical procedures and images and scans. And so something like last night happens, we're clearly excited because it's been very difficult to not -- for the last whatever months to not tell you what's going on.

But it's not really about us, it's about them. And it's their story, it's their vision. We're just creating and allowing them to have the core technology to help their dreams come true. So what -- so Midjourney, as many people know, is a text-to-image company. They take basically if you think about it, a very little amount of data and they extrapolate into something extraordinary. And they've proven to have incredible AI capability and an amazingly creative group as well. And it's just been a privilege to work with them on their vision.

And there is something called tomography, and I think that's an important basis of the foundation, it is -- like every ultrasound device is able to both send out an ultrasound signal, but it also then listens to that signal. So like our handheld ultrasound device, you're able to send sound out that sound will then bounce off of anatomy and how that sound comes back, the ultrasound device listens then -- after it sends a sound, it dynamically listens and then that produces an image.

Well, tomography is different. Tomography uses the same send and receive, but the same device doesn't send and receive. You put one device, say, at 12:00 on the dial and then one device at 6:00 on the dial and you put the human body between them at the center of the clock. And you can send a sound wave out from one device or one chip, and it will actually -- as it interferes and interfaces with the different types of tissue and bone, different sound is created and then listen to at the other side of the dial. And so I'm going to show you, first of all, I'm going to -- for everyone, Josh, do you want me to go through and show some of the videos that everyone...

Sure, it would be great.

But this company, Midjourney, who has this massive compute and creativity wanted -- has a desire to bring imaging and to bring their capabilities to help humanity. What they want is the ability for people to have the opportunity to have an MRI like image with no consequence. So no radiation and no magnetic consequence at incredibly low cost. So almost like going to the gym, going to the spa in your normal course and at the end of it, just getting an image at an extremely low cost.

And then that image is a way in the beginning for you to be able to just monitor your life. How is my BMI as I'm working out? Which are my muscle groups that I'm working on getting bigger? Are they -- is my body responding in the same way? If I'm eating different foods, how does my body respond? So there's a whole wellness layer that is pre-FDA. And we see all the devices like Fitbit and Whoop and all the other capabilities of looking at the body.

Those are all in scope in the very beginning. But then over time, codifying an FDA approval -- we are an FDA-approved company. We have done tens of millions of scans. And our platform, our technology is all FDA approved. And again, they stand on our shoulders, and we are delighted to help them bring this vision to fruition. So I'm going to navigate here and then try to show you the initial video that showed the concept, and then we'll show you a video that also shows you how the technology works. So I'm going to share sound. I'm going to share the video. And if you could just tell me, can you see that video?

That's coming through, yes.

All right. Well, so here we go.

[Presentation]

There is that first video.

Thank you for that, Joe. And that's a nice foundation to build on -- build out our discussion from. Maybe just take us back to the early days of this partnership with Midjourney. I believe it's been about 18 months where they used the Butterfly Ultrasound-on-Chip semiconductor technology and started this development program.

I think one of the questions we've gotten since last night is just how not being in the imaging world -- medical imaging world prior, it seemed like a pretty rapid development process and how -- I apologize for my video difficulties there, but how they've been able to accelerate so far? I mean, clearly, your Butterfly technology has facilitated it. But -- and where does it stand today or maybe next steps in terms of getting it into the commercial era?

So for those -- for anyone who's a student of ultrasound, the concept of tomography is not a brand-new concept. And there is very strong research out of Caltech and out of other universities around the use of -- the capabilities and the potential for tomography. And one of the key things that Midjourney and what David's vision was the ability of bringing these things to life with technology. They -- the beauty of this is if you look at the intellectual property of Butterfly, we've thought about this a long time ago.

We have very deep IP that specifically calls out tomography and had in -- the early vision of the company had contemplated tomography to be done. What's really specific about Butterfly is we can tune the chip. And so we can -- when we talk to partners, we bring the chip into a laboratory, they ask us if we can do certain things, we modify the software and it sends a signal out right then and there. And so the ability to go from vision that David had to an actual prototype is an extraordinarily compressed amount of time because all of their efforts are focused on the algorithms of the collection of sound and then the ability to process that.

It's not in the foundational work that's got to be done on how do we deliver the sound and how do we get -- it's because we are infinitely tunable, they walk in day 1 with a vision and then they are tuning and tuning and developing and tuning and refining to their vision, which allows the development to go so quickly. So they don't have to burn cycles on the core development of the ultrasound itself from a capability standpoint.

So we set up when they, like our other partners, purchased access to our SDK, purchased chips, asked us to help them with certain things and then they were off. And they did all their work and didn't have to replicate the work that we did. They started where we left off, and we gave them everything and then you can see their extraordinary vision came to life. And that's what's so important. And I think for a Butterfly investor to know that we have a platform that is so fungible that can be -- that we can give people a launching place to then just accentuate their idea and then make something of it is why Butterfly has a massive TAM expansion.

We're not just the TAM of point-of-care ultrasound anymore, but that's what we are developing. We are walking down that trailblazing path of handheld ultrasound. But now our investors can participate in so many new markets because we can allow people to start where we left off. And so we are completely delighted. Our -- all 200 of our Butterfly employees are delighted, and we give our full support to help David and the team and Thomas and the team at Midjourney realize their dreams. And we're going to be doing it for a lot of other people, too, in a noncompetitive way.

Appreciate that. And maybe just to talk about Butterfly shareholders and Butterfly company on this -- with this Butterfly Embedded franchise, including the Midjourney partnership, positively impacting shareholder returns and financial metrics. It would be great to just review the agreement with Midjourney -- I mean, $74 million of expected payments over a 5-year term and just I mean, how we think about this partnership from here impacting the P&L from revenue down.

Sure, Josh. Can you hear me?

Yes.

So as we've talked about in the past, the agreement has pretty much 5 different parts of it, which we can get our revenue from. There's the upfront tech access fee, there's an annual licensing fee, co-developing work, chip sales and revenue share. Certainly, starting in fourth quarter of 2025, we started realizing revenue from the first 3 of those for the most part, the tech access, annual license and co-development. There was also some small revenue recognition from chip sales over time. And we've been working closely with Midjourney, but we're still working through this as well.

Midjourney is evaluating their potential business models, for example, whether it's a subscription-based model, monthly or annual, whether it's a charge per scan model. So there's still things that are moving around and we will continue to. Even David mentioned various options they were considering last night, and it's not just locked in. They're still -- as they roll out from 2027 and beyond, they're going to be learning as they should be.

Our initial assessment, which is just that initial and is not specific guidance. It really obviously results in a large potential opportunity for us once they are commercial, and really from 2029 and beyond. And for the most part, that's where there'll still be revenue from the other components I talked to. But this is where we'll start to see an uplift in chip sales as well as the revenue share component. And if I just reference kind of Midjourney's target of 50,000 scanners in the market by 2031 and certain assumptions that we made around this rollout, this has the potential to result in between 1 billion to 1.5 billion for Butterfly by 2030.

And that's with increased contribution coming from, as I mentioned, the chip sales and the revenue shares. At Midjourney's sales, we do anticipate more than 50% of our Embedded revenue would come from the sale of chips to Midjourney. So it will be more chip dominant than revenue share dominant, but both contributing a significant amount of revenue above beyond where we are today, of course.

In addition, when I translate this into our longer-term plan, we are taking a disciplined approach. But even at a quarter of these levels, total revenue, so core plus Embedded would be above $550 million by 2030 with Embedded obviously representing a more significant piece of our revenue with core still growing at a solid high teens that compound annual growth rate.

So in taking a conservative approach to how we're modeling this in no way shape or form, it kind of speaks to our belief in what Midjourney is going to accomplish. Incredibly talented group of folks. If people tuned in last night, they certainly saw that effective there. We're familiar with it. They are as well, and we're very confident in what's going to happen here.

Well, I appreciate that. In your press release today, you called out 40 Butterfly Ultrasound-on-Chip imaging modules per system. I mean I think we've gotten some questions around just the unit economics to Butterfly per scanner. Is there any way to -- just help us understand the 40 Butterfly Ultrasound-on-Chip imaging modules? And is that 40 chips or semiconductor chips? Or is a module more than 1 chip? And I think you broke down some of higher level eventual revenue contributions.

In the specifics -- but we do expect that as a number that will increase, certainly a step-up as they move on to their next -- we move on to the next-generation chips and they move on to the next generation of their scanner.

And I know Midjourney detailed some of their timelines for development moving from prototype to research model to commercial model last night at the event and I think your team shared a link to a website where Midjourney lays this out. But maybe we can just review that for some of the investors here just in terms of those timelines that you referenced, John and moving from current prototype and unveiling to research module to commercial module and some of the steps there and I guess their confidence level that those time lines are realistic.

So just to frame it. So if you listened to the talk last night, they said the goal was to have a full body scan within 60 to 90 seconds. That's a massive amount of data. David can walk you through, and we have a tech video here, and I'll ask David to talk through that tech video in a bit to give us some transparency. But currently, with our 4.3 chip, which is the chip that's in this -- the scanner that you see, which is the only chip that we have any type of volume on today, that chip would probably take 20 minutes to do a full body scan, as David also referenced last night.

There's then an architecture in a second generation between chip and Ethernet that gets it down to 8 minutes. And then it's our Apollo chip that -- I've mentioned to everyone many times that our Apollo architecture will have 20x the processing power of our current chip, which then means that Apollo will be able to acquire enough data and send enough data through to get that scan down to a minute.

So when we talk about first, second and third generation and when David last night mentioned that there's a custom silicon chip coming, we've contemplated Midjourney's requirements in the development of our next generation -- of our fifth generation chip, which is our Apollo platform. As you recall, our fourth-generation chip is our Harmonics or MEMS-based platform, and that is in development or in production right now with TSMC and sometime early next year, our next probe will come out with 5.1 and then Apollo.

So Apollo is key. And Apollo is a 2029 event. And when Apollo comes out that's when Midjourney is looking to scale, and that is kind of let's -- if you want to call it the limiting factor or the catalyzing event is, Apollo is not only going to be a major impact for point-of-care ultrasound and be the fifth generation of semiconductor for point-of-care ultrasound. It is a major leap forward in technology that's going to allow image processing at tremendous speeds.

And several other of our partners that you'll learn about in the future are also waiting for Apollo for a lot of different reasons, but mainly that processing power. So consistent with how I've communicated to investors on our evolution of semiconductor is consistent on how Midjourney and also other partners are thinking about the future.

Thanks, Joe. I need a little compute power on my side to help the video issue I'm having. But...

I'm not complementing you on your looks anymore because clearly, someone didn't like that.

Well, maybe a follow-up before we dive deeper into the tech with David is just -- I think you referenced earlier in your initial remarks, Joe, I think I heard FDA, but just this initial venture looks like in the consumer health and wellness segment that Midjourney is pursuing. And then ultimately, is it a goal to have an FDA-cleared full body ultrasound computational tomography platform?

Yes, yes. So I think going for wellness allows for Midjourney to get started now and get started with scans and get the learning process going. And there's no regulatory requirement because there's no clinical diagnosis. This is just taking images and there's -- no one's pointing to any type of consequential clinical outcome. But David got questions last night around FDA, and you can see the tremendous respect David has for FDA, which we all do.

But again, as I mentioned, Midjourney is standing on top of Butterfly's shoulders. When it comes to our tech and our capabilities and everything we have is FDA approved and validated. So what needs to be validated purely are the algorithms and how the images is extrapolated and what claims they're going to make. But so much of the foundational work that would normally have to be done, we've already done it. and we'll be with them all along the way.

So the confidence level on our side is very high that they'll get the claims and get the clearances because this path has already been blazed and their unique capabilities. I think -- just think about it, just think about going to a spa or going to a health club and once a week, having a scan and getting to see what your BMI is, getting to see where fat is building, where muscle is building, getting to see the pulsatility and the elasticity of your vascular system.

And it's just absolutely amazing because we always hear stories of finding things that we never knew. And there's also so many disease states that are asymptomatic, like pancreatic cancer. You hear every -- you hear -- if someone gets pancreatic cancer, it's a tragedy because they're not symptomatic until Stage 4. Lung cancer. They don't -- unless they're lucky to get a scan that it's just such an asymptomatic disease that when symptoms do present themselves, it's almost too late. And the patient then has to go through extraordinary interventions in order to save their life.

But now if you can have this as a part of your every day, if you can have this as a part of your every week and your journey, then you can learn things when they happen really, really early. You can learn things when they have an opportunity to be intervened. And that changes the world. We have a sick care system today that you can't go get an MRI to determine how you're feeling. That's nuts. You can't get that because you can't get a prior auth from -- you won't get a prescription from a doctor. You won't get a prior auth from a payer.

And then you're in line at the hospital. So it's -- the MRIs are reserved for sick people. It's not preserved for -- or an injured person. But to have this type of imaging as a part of your everyday life and have it be low cost, have it be no artifact, no radiation is an extraordinary change for humanity. And that's what Midjourney is set out to do.

Excellent. And just a follow-up on the trajectory that Midjourney laid out and that John referenced is 50,000 scanners deployed in 2031. Does that include that mostly in this consumer health and wellness segment? Does that include potentially in care settings, medical care settings, hospitals, imaging centers, et cetera?

Well, I think, again, based upon what David said last night, their vision is they have the ability to build this out themselves. And with the capital that Midjourney has and then potentially future investors in Midjourney, I think he put the CapEx number on the 50,000 scanners at about $20 billion. And he felt like as an organization, they had the ability themselves and with investors to be able to do that themselves.

So I don't think initially, there is this desire to jump into the center of the health care system and to get into -- and given the cost, the type of volume they can do and the type of cost that they can provide, I don't think they're really looking at getting into the barriers of will this be reimbursed or not because the cost is going to be so low, it could be out of pocket. And I think their vision is one of controlling the distribution, controlling the narrative instead of going into a system with a lot of roadblocks.

So yes, I think they want to get the FDA clearances, so they are not having any claims to consumers that are untrue or misleading. But I don't think that -- I think they'll be reserving the existing imaging modalities for patients who are entering the sick care system and the new health care system of the future where people are monitoring and they're taking accountability for their behavior and they're learning and growing.

That's really the -- if you want to identify what's going to fix our health care system, it's when patients are aware or people are aware of how their behaviors impact their health. And this imaging modality is one of the ways that you can really have awareness of changing your behaviors and being more in control of your own health journey.

Understood. And it sounds like Midjourney Medical's mission is to democratize access to full body scanning, which is very consistent with Butterfly founders' kind of mission to initially -- with handheld ultrasound to democratize access to advanced imaging. But maybe we can dig a little bit deeper on the technology and the prototype that's in play today and love to just better understand the platform for one and understand where this kind of prototype sits and how, I guess, there's some level of derisking that has taken place in order for Midjourney to be able to unveil the prototype and love to just learn more about that, too.

All right. We're going to put the video up.

So the core of our technology is what are called CMUT, Capacitive Micromachined Ultrasonic Transducers, and each one of them is a tiny little membrane like a vibrating drone. In this image, you can see one of them...

[Presentation]

Surrounded by others nearby. And what they're doing is they're vibrating. And as they vibrate in a fluid, they're launching an ultrasound wave into the fluid. Can you notice, this is a 2-dimensional array of these devices. Each one is about twice the diameter of a human hair. So they're tiny. There's about 9,000 of them on a chip. And the fact that it's a 2-dimensional array is really significant. It's part of the differentiating capability of our technology compared to other ultrasound systems that are using older piezotronic-based transcers.

Our chip is really a 2-dimensional array, and you get -- as we scale back, you get a sense of the scale of 9,000 of these pixels, ultrasound pixels on a chip. So just imagine this as a little bit like a camera chip. And our competitors have one-dimensional arrays. So where we have the 9,000 elements in a 2D array, enabling you to capture a 3-dimensional image, our competition has 128 elements in a linear array.

So it's -- I mean, the analogy isn't perfect, but it's like we have a digital camera like your iPhone camera and the competition is still using film. In the Midjourney application, you can see there's 40 of these arranged in a ring, and they're operating together to make an entire system. And if you look also -- you might imagine or wonder what's sticking out of the -- outside of those rings, those are actually our probes, basically the electronics from our handheld probe is what's used to build the prototype.

And you'll notice that inside the ring, you can see illustrated ultrasound waves transmitted from one device and then being received by multiple devices on the other side of the ring. That's the tomography aspect of it. And as the waves pass through tissue, then you can see what the different layers of tissue are, the bone, muscle, will have all the organs. And maybe a key thing that some people might be wondering is how does Midjourney identify what the different organs are.

In some of the images, you might see that they're colorized and show you what the different tissue is. That's because there's differences in the speed of sound in the tissue. And so you can actually identify the difference between fat, muscle or organ just based on the time it takes for the ultrasound to pass through the system.

And in this -- I guess maybe the other thing to highlight here is people ask how is Midjourney able to build this thing. So as Joe said, we started with a semiconductor chip, but we have a software development kit in SDK that allows you to use the chip in lots of different ways. So we have a lot of capabilities in the chip that we aren't necessarily using in POCUS today because POCUS is a handheld device. There's only so much compute power that you're going to put inside that handheld device.

And so with partners like Midjourney, what our software does is it allows them to unlock the capabilities of the silicon that are in there to enable whole new imaging modalities that really don't exist today. And this is just an example of one kind of partner that's using our technology to really enter a new vertical for us to enable a new adjacent space. In this case, it happens to be medical imaging. We have partners that are doing things that are even further out on this, but again, enabled by the same silicon and by the same SDK.

That's impressive. And just thinking about -- I'm not an engineer, so I don't know if my opinion matters as much. But just thinking about where the prototype sits? I think, Joe, you mentioned that the speed of the scan will -- is dependent on next-generation chip technology getting out to Apollo to get to that 60 to 90-second full body scan timing. But I mean, where are we or the journey in terms of scanning humans and capturing these images that look very similar to CT scan images and even MRI images to a degree. Is it [indiscernible] image or is there an output that's already in play? And is this video representative of human scans?

Yes. So they have -- I think last night, they mentioned they've scanned a dozen people so far, and then they're scanning phantoms and doing a lot of foundational work. And a lot of this is really early and preliminary. But remember, what Midjourney has done masterfully is take a little amount of data and be able to learn from it and extrapolate it into something extraordinary. Just go on their website and see what they do.

The reason why they are the best text-to-image company is because of the marvelous creativity and intuition on the ability of understanding something and representing it in an extraordinary way. So the ability to take the black and white dots of an ultrasound and learn what they actually mean and then create a format that allows you to understand what it means is something I think we're going to see that incredible capability of theirs really shine and really make a difference.

So they mentioned that their first spa will be open to the general public by the end of '27. And so I think that will be the second generation that they're working on. And I think they'll be just collecting as much experience possible through that time to refine the image, refine the image, refine the AI around the image on how to interpret what this means and work with the medical community to ensure that they understand exactly how this all should be correlated.

And by that time, have their FDA clearances and have their freedom to operate when Apollo comes out and they get into '29 and really hit and really go fast. I think that's kind of the whole thought process. And by that time, having a processor with the speed that we have and also they're using some of the latest NVIDIA technology and Holoscan and when all this comes together, it will develop truly an extraordinary outcome for people.

Midjourney is doing something magnificent for humanity. And this is not an understatement. We have big imaging who kind of captures inside the sick care system, but this is truly an opportunity to get outside and get earlier to intervene for patients. And it takes an innovator like David to make this happen.

Awesome. And just a follow-up on, you mentioned big imaging, the major imaging companies with manufacturers of CTs, MRIs, PET scans and all the big capital advanced imaging technologies. I mean have there been previous development programs to try and develop an ultrasound-based computational tomography platform in the past?

I mean with Butterfly's Ultrasound-on-Chip technology, semiconductor platform, compute power that's accelerating as you move into next-generation chips seems to have unlocked this. But have there been attempts in the past to use ultrasound for computational tomography? Or is this really the first attempt to kind of create a whole new imaging technology platform?

No, I think from what I understand, that tomography had been studied for years and had -- and there have been some companies even up to recently, who have developed tomography-based solutions, but they've been built on a piezoelectric platform that is -- you have to live with the frequency that you own. You don't -- for example, there's 3 different ways of scanning -- 3 different ways of the core ultrasound arrays are there's linear curve, linear and phased array. Those are 3 completely different arrays of sound.

And with our chip, we can dynamically program and we can send out arrays to the specific requirements that the companies have. And I think that the missing link to really make this commercial has been the dynamicism of our chip and then the software capabilities to rapidly innovate. But yes, this has been -- if you do search it, you're going to see a lot of stuff that's out there. But I think there's such a focus on the highest end of imaging in the health care system that it's taken a GenAI company with mass compute and tremendous AI capabilities to reimagine and disrupt the space, and that's what Midjourney is doing.

And has Midjourney received any feedback from the radiology KOL community just in these early days, and it is an ambitious journey that they're taking on in partnership with Butterfly, but just curious if there's any early buzz being...

Absolutely. They have a host of advisers that are of help. They're a very smart company. They've been thinking through their regulatory strategy. They've been thinking through their clinical strategy, and they have access to some great minds. So absolutely. And then to whatever extent they need us, we're here for them, but they have tremendous capabilities and resources to get the information they need.

Excellent. And I appreciate you guys wanting us to go deep in spending all this time on Butterfly Embedded. I'm going to continue to go a little bit deeper just with this kind of unveiling with the Midjourney collaboration, partnership and agreement. My understanding is there's a pipeline of potential partners and would love to just regroup there and if there's any updates in terms of talking about that pipeline and how we should be thinking about additional partnerships being signed or being made public as we move through the course of '26 and into '27?

You'll see more. I can't really -- I can't talk about anything until they're public, but there are just some extraordinarily creative people who I think we've tapped into a nerve here where a lot of ideas that have existed in the history of imaging are now being able to be brought to fruition with the fluidity and the ease of use of a programmable chip. I mean before, it's just such a big research program to have to go through and make modifications in a solid-state fashion.

But we're talking to some extraordinary people, Josh. And the partners that we've already signed are extraordinary and some of them want to tell their story. And when they do, we'll be very excited about you hearing their story. And then there's a lot of new people that we're having conversations with that are -- that have David Horsley like creativity and ingenuity. So you're going to see other things.

This isn't a one-trick pony. It's just Midjourney had a heart attack when we signed our deal, and we told them that this was material to -- we're a public company, and this is material, and we had to go and file our 8-K and let our investors know they just weren't ready for this type of scrutiny yet because they still very, very are. Other partners also feel the same way, and they're in research right now.

So I know the investment community wants to see what's next, wants to see everything that's happening. I told you this was happening. I told you what we were doing with our chip. I showed you the different markets that we were talking to people in, in our slides with the big circle with the bubbles around it. I told you we were going to sign a deal like a Midjourney deal. I told you Midjourney was going to be big and creative. And I'm telling you there's more.

I'd like you continuing telling us these positive things, Joe. Thank you. I wanted to circle back, I apologize to Midjourney. Just one of the questions I had was some of the comments that Midjourney's leadership shared around Midjourney scanner being 1 of, I think, 4 platforms, 2 potentially being handheld and another, I guess, bigger capital systems, similar size potentially to Midjourney scanner. But any comments there on the build-out of Midjourney Medical's hardware portfolio?

No comment.

Okay. Understood. And back to Butterfly Embedded. I mean, Forest Neurotech was, I think, a partner through Butterfly Garden. There's been some, I think, collaboration or partnership or some of -- that team has moved over to Merge Labs. I mean is Merge Labs -- does your partnership or collaboration with Forest move over to Merge? Or is there anything of note that we should understand between Butterfly and Merge?

We are very excited about our relationship with Forest. Forest is like an incubation laboratory where also they have some of the smartest minds that are coming up with some of the biggest things for the future. And ultrasound is a big part of the expertise that they've established. And so we're very excited about the partnership and the progeny that's going to be coming from that partnership like the Merge.

We hope that there is a future relationship. And if there is, we'll talk about it. But Merge themselves are very, very early, and they're doing a lot of investigation. And if something comes to fruition, we'll let you guys know. But right now, we're focusing on our current partners, and we're focusing on the relationship with Forest and a few other BCI companies, too. But they're doing some extraordinary things over at Merge, and they're just a great team of people.

Appreciate that answer. And just thinking about other utilization of ultrasound outside of diagnosis, uses focus ultrasound as a treatment as an intervention, there's potentially some brain computer interface technology development out there that could potentially use ultrasound. And maybe on those 2 topics, ultrasound-based therapy and brain-computer interface, any just high-level thoughts to share in terms of how Butterfly is positioned to help in development programs in those 2 vectors?

Yes. So David runs our business. He's having a lot of these conversations. Don't tell them too much, please, but give them a little bit of a perspective of those type of areas that we're working on.

So yes, BCI, brain computer interfaces, are one of the other verticals you'll be hearing more. And obviously, you already know about for us. We have a number of partners that we're working with there. As Joe said, a lot of them are very early research, but there is kind of an emerging consensus that for noninvasive interfacing to the brain, even transcranial outside the skull, no implants, that ultrasound is really -- has some unique capabilities.

And that goes -- so you'll, I think, begin to see some really amazing brain imaging data that may come from some partners. And there's also some people, as you mentioned, the focus ultrasound versus stimulation, so you can potentially both read and write. So that's an area of BCI that is definitely something we're very excited about. It's a targeted area for us, but it's just one of many. So there are -- in addition, we have partners that are much more adjacent to what our core focus business is.

So they might be using our probe actually but just in a different way, incorporating the system, whether that's a surgical robotic system or other things. So we are really -- we are looking at a very wide variety of different potential markets for the technology. And the key thing is that it all leverages the same chip and the same software.

So we developed a core technology, a little bit like the way NVIDIA developed GPUs and CUDA for starting with one application, which is for gaming and expand it to so many other things that we think that ultrasound and the chip we developed it for POCUS, but we're really excited about so many other areas that technology can be used today. And that's really what Butterfly Embedded is about.

And maybe just a couple of minutes left, Joe, I'd love to just have you review as you've already done, but just maybe just a little bit more detail and time lines around next-generation chip development and those programs, how derisked they are? I know you're making progress for one. Maybe we'll start there. I just have one follow-up.

So hopefully, this time next year, maybe a little sooner than this time next year, we'll have our next-generation chip, our Harmonics chip on the market. And the chip that has the Harmonics capability will be another leap in image quality. It will get us even deeper into subspecialties. It will be -- in my view, I actually scratch my head as to why after this chip comes out, anyone would buy another handheld ultrasound device because the whole argument in the past was, well, this one can do a better cardiac image or this one can do a better OB image, although we do everything that they can't do. We should be able to do those better.

And when that happens, what's -- why would you buy a single-dimension handheld ultrasound device when you can have a full body device that has better imaging from top to bottom. So we think that the market is going to change a whole other step function with that next chip and with that next release and we're extraordinarily excited about it.

After that, we've talked about IQ Station, which is our entry into the cart market. There's billions of dollars of revenue sitting in hospitals for carts, and they're having to buy all of these unintegrated systems. And we have a complete vision how hospitals will work in the future that blends the one-to-one model with deep clinical in-person environment. And that is something that is also a 2027 occurrence, probably a little bit closer to the end of the year, and it will build on us having that next level of image quality.

But the software that stands behind all of that, that connects it and the workflow is going to be brilliant. We're not just going to be launching a device. We're going to be launching an entire solution that integrates into the way -- into practice of medicine in a way that's going to make things completely different than the current market. And so we're very excited about our core focus business.

And we -- and all of those innovations and everything that we do with our software and our chip and all the learning and imaging, all of that translates to our Embedded partners, too. This is an incredibly symbiotic system that as we get better and better, we then help our partners get better and better. And then we learn from our partners. And then -- and it becomes this wonderful cycle. And then we get into the 2029 time frame when Apollo is available.

And I think that's when either then or before some of our other partnerships really rear their head because they're kind of -- some of them are really waiting for that processing power and you'll see what that means. And then, of course, Midjourney will be stretching their legs at that point in time, and it will be a very powerful time for Butterfly.

And so -- and our team is even starting to contemplate what's after Apollo or what's the next versions of Apollo too. So we are not going to let go of our pathway of chip and semiconductor development. And that's where you're going to see significant separation between where we bring imaging and where it sits in the legacy health care system today.

Awesome. And maybe last topic or question. I think on the third quarter call last year, you referenced that Butterfly was point-of-care ultrasound company with a semiconductor business unit and ultimately, it could evolve to become a semiconductor company with a point-of-care ultrasound business unit. You're on your way in that direction.

And I think one question we get is just on the IP portfolio, the protection, the moat you have on the IP side, but then also just on the secret sauce and know-how side with your team and the partnership with the Taiwanese semiconductor manufacturing company. Maybe just hit on those 2 moats and then that will be a good place to end.

Well, I think if you went and searched semiconductor-based ultrasound, you'd see a lot of research programs, you'd see universities, you'd see other companies who've done work and you'd see copies of the chip. There is a very big difference between coming up with something on the table and then getting it commercial and getting it at scale. And so much of our IP is not even in the IP that we have filed, but it's in the know-how of how to build the supply chain. It's a know-how of how to get this into production.

And a lot of that rests between ourselves and TSMC and some of our other partners where we've gone through a significant learning curve. Our first chip cost us $300 million. So -- and in that $300 million, we learned like 20 different ways not to make a chip. And we learned all the different things that don't work. And that's not sitting out in our IP. So anyone who takes a step down this path is going to have to walk down that path. There's no one -- we certainly aren't going to make it easier for anybody to share what we've had to go through to get to this place.

And it took our founder and it took the initial CEOs that were my predecessors, a lot of fortitude to get through that pathway. But investors believed in the vision. We, of course, nothing ever happens fast enough. I think you go back to a lot of the great tech companies that are public today and follow their stock price all the way to the very beginning.

Look at Apple at the very beginning. It's just this little thing for so long and then the chart explodes because it's -- this is so hard to do. And someone is just not going to come in and say, "Oh, well, I've decided I want to get into this business. Yes, okay, we'll see in 5 years." And we'll see after $300 million and your discovery. So I do think that we are on to something. And I think that the vision of the builders of this business was premonition. And I think our investors are going to benefit from that for a while.

Outstanding. Well, I can't thank you enough, Joe, John and David for letting us participate in this big week for Butterfly, the Midjourney scanner and Midjourney Medical unveiling last night, and I appreciate you sharing all the details today and taking on all these questions. Exciting times for the company and for the imaging space.

Thank you for your support.

All right. Thanks, gentlemen. Have a great rest of the week and weekend.

Good morning. Thanks for joining us here on the final day of the William Blair Growth Stock Conference. If you don't know me, my name is Andrew Brackmann. I'm the diagnostics analyst here at the firm. Very happy to have the team from Butterfly joining us. We have the CEO, Joe DeVivo; and SVP of Finance, Megan Carlson, joining us here. We'll do about 30 minutes of presentation in this room, and then we'll head to the breakout in Burnham A afterwards. And then lastly, for a full list of research disclosures, please visit williamblair.com. With that, I'll turn it over to Joe.

Awesome. Thank you, Andrew, and thank you to the organizers of the conference. Thank you so much for having us here. So growth conference. So generalist investors in general, we've -- Butterfly is a semiconductor-based ultrasound company. Now, what does that mean? That means that in existing ultrasound. So everyone know what an ultrasound machine looks like. You've seen them before. So you have the overall machine and then you have the different handles and probes. And each of those probes are designed to see different parts of the body. Our founder in 2011 endeavored to build a chip. And that chip instead of having multiple probes that use a crystal lens, that chip would be designed to deliver ultrasound. And it's the first semiconductor ultrasound in the world.

Now from 1993 to 2005, this little company called NVIDIA was a video game company, right? You all remember that. They made chips for video games. They didn't make chips for data centers. They didn't have on their plan the desire to create this for that market. But what they did was they solved a foundational problem for themselves. They went out and built a chip that had the processing to do multidimensional calculation to do very sophisticated stuff and then even more chips to allow other gamers to come into this environment. And this happens every once in a while, where a company solves a problem that they have and then realize that everyone else has a similar problem.

You all saw that picture of Jeff Bezos sitting at his -- I think it was a folding table desk with his old CRT computer with the handwritten Amazon.com on it. Well, when he was at that table, he didn't have a vision to build cloud compute. They got to a point where their business grew so much that they got to a point where they realized that in order to do online retail, they needed to have a global cloud infrastructure. And so they spent billions of dollars building a cloud infrastructure for themselves. And when that infrastructure was completed, they were only using 1% of it because they had to have all the fixed capabilities of it. And what did they decide to do? They decided to sell capacity. And they became -- that became AWS, which is the most profitable contributor to Amazon today.

So my point is that sometimes you solve problems that are bigger than the problem that you were trying to solve. And I'll maintain to you today that Butterfly has done that. And what we've learned over the last several years is that the semiconductor that we've developed for ourselves is very valuable to other people and has many applications. And that semiconductor now, we partner with TSMC. TSMC builds our chips. They help us in development. We have over 600 patents in this space. And we have with MEMS now and with the ability to control all of our sensors have a tremendous ability, and we're learning with the amount of inbounds we're getting that other companies would like to have access to this technology.

So we start off in our first application, the reason why we want the chip to do everything is because a lot of times, doctors encounter patients and they want to be able to scan the entire body. Patients come in the emergency room, you don't know what's wrong with them. Is it they have pump fluid in the lungs, do they have a bowel obstruction, do they have an appendicitis? Is it a swollen ovary or endometriosis? Is it whatever? And they -- the only way you make those diagnoses today is if you either order an ultrasound or you have the ability yourself to do an ultrasound and you're present with a machine. So in general, if you can have one device that has the capability of seeing the entire body, you now have a very, very powerful device.

And today, when you just look at the medical need, 2/3 of the world doesn't have access to medical imaging. They don't have a nice radiology clinic to go in and sit with an ultrasonographer in rural areas and also in remote areas. So on our primary -- the business that we're developing organically, we're solving a major problem by having the semiconductor because they plug it into their phone and have the ability to select pulmonary, cardiac, MSK. And when they select a setting, the setting is automatically changed to deliver the array and the frequency that, that clinical indication requires. So that initial need, we've now sold over 150,000 devices around the world. And we see that there is going to be a real significant uptake in the future of what's called point-of-care ultrasound.

And just to show you how it works, you see on the left that animation that there's -- it's a little vibrating drum. And I like to think of that as a pixel on a TV screen, right? So that one drum can create sound in a certain way. But our chip has the capability to control all 9,000 drums individually, right? So when we talk about us being a health care company today, it's like asking yourself, well, what type of programs do your TV -- are they compatible with? Do you have a country-western TV? Do you have a TV that plays dramas? I say, "Well, what are you talking about?" It plays all shows, right? Because each of those pixels are controlled individually. You change a color, change the color of all your pixels, now you have a picture.

Well, with Butterfly, we -- when you can control all 9,000 sensors individually, you can not only meet the 3 clinical use cases that you see up top being curvilinear, phased array and linear. And we can also now do basically anything. Again, ask me what program I can play on a TV. I'll tell you everything. Well, what can we do and what type of an array can we put out? Well, I can control all 9,000 sensors. So we can do biplane, we can do 3D. We can actually -- because it's a semiconductor, we can literally move the beam on the device without moving the device itself, where a normal ultrasound device is like a flashlight, you have to contort it.

So that is a very, very powerful set of circumstances. And historically, supercomputers used to be huge. Now every one of you have is one in your pocket. Ultrasound machines used to be huge and then the power of processing, every doctor and every nurse in the world will have one. And so we're very excited about what we have done in our core business and where our core business will grow. And we're just as excited about a lot of new opportunities that are coming our way because we are now starting to license our chip. In 2006, NVIDIA created a developer software called CUDA. And that developer software was the first time that a third party can get inside the GPU of an NVIDIA chip, change the settings for its application. Well, we've replicated that. We now have software that allows third parties to come in and modify our chips. And I'll talk to you near the end of the talk about the types of companies that we've already partnered with and that we're going to partner with.

But Butterfly, we sell directly to the doctors. We sell to ambulance companies, clinics, cruise lines and of course, hospitals. And as we built a brand, we've become pretty popular in the medical community. And just recently, over the last 9 months, there are 3 different TV shows that have featured Butterfly, and we didn't place this product. Organically, there was an adviser on The Pitt who is a big Butterfly point-of-care supporter, and they had about 6 different scenes where they showed using the Butterfly.

Season 50 of Survivor just started and the inaugural episode, one of the kids -- one of the contestants, some old. One of the contestants hurt their Achilles and they right on live TV -- not live TV, but right there on the show, did pull the Butterfly out. And then on Virgin River, which is my wife's favorite show, I've never seen an episode, but they have a scene where the patient's [ wife ] has lower right quadrant pain and she's diagnosing an appendectomy. But Butterfly is now, especially in the medical community, becoming a verb. Did you Butterfly that patient? And it's really cool to have a brand that is popular amongst the doctors and nurses that is understood. And there's -- we have example after example of that type of overall acceptance.

So we're the only -- I think there's one software company. But aside from that one software company, we're the only medical company in the world to actually receive an Apple design award. I have a big like titanium cube on my desk. Very proud of this award because we're architected in a very modern way. Our design is very simple, artistically done and very powerful and fungible, which is very consistent with the Apple brand. When I used to run other medical companies in the past, and I had a piece of hardware in the hospital, if I wanted to update that hardware, I'd have to send a rep in the hospital with a USB stick, right? Well, actually, today, we -- all 150,000 of our devices that we've sold have the ability to connect into our cloud. And just like your iPhone, every picture you take gets replicated in the cloud, right? And if you want that picture, then you have to sync up with the cloud. Well, every image that every doctor takes in the world comes in our cloud.

We have over 25 million images today, and that is growing at just about 30,000 to 40,000 new images per day. So with 25 million images, growing 30,000 to 40,000 a day. That becomes a very potent AI warehouse to be able to develop new AI applications because you can imagine a developer of AI ultrasound coming up with a model. And now all of a sudden, they need 5,000 kidneys to train their model. Well, where do you think they get those 5,000 kidney scans? They're not open sourced. And we don't sell that data. But I'm saying they'd have to go do a clinical trial. They'd have to go to some university and do all these scans at very, very high cost, very complex. We -- if we're developing a new renal model, we can just simply pull in the anonymized image and build the model right then and there. So we have -- the value of this is powerful.

But we also -- what's happened is we've now -- as our popularity has increased on the medical ultrasound side, we're getting -- it's creating a little bit of an issue for hospitals because now everyone is walking in the hospital with a Butterfly and it's not linked into their EMR. It's not linked into their system, and I get a phone call from the administrator yelling at me for selling all these doctors. But we had one major hospital that you all know was upset, and we went back and realized that in their town, 140 of their doctors had bought a Butterfly.

So we built an enterprise software system called Compass. And that enterprise software system takes all the data from all of those devices and pushes it into their EMR of choice, pushes the image into their DICOM and pushes a file into their revenue cycle management system so that when they're acquiring an image, today, hospitals that are using point-of-care ultrasound only capture 35% -- 30% to 35% of the images. So it's just so easy to pick the device up, record an image and then put it down to pause and do all the documentation takes time. And so it's a revenue loser. And also, it's not compliant. You're required as a hospital to have all the data for that patient put into a record.

So when someone uses our enterprise software system, they're able to acquire. And we've now seen with our installations of probably about a couple of hundred sites across the U.S. health systems across the U.S. have it. And when they put the system in, they get to about 70% to 80% compliant. So it allows them to double their ultrasound scanning revenue the moment they put our software in place. And then lastly, we've built out just like thinking in the ecosystem and platform side, we've compartmentalized all of our capabilities to be able to link into not only the hospital's network, but to allow ourselves to link into third parties.

So if any of you have seen ultrasound, ultrasound is hard. It's easy -- if you look at an MRI technician, their biggest worry actually is that patients got metal in them because they're concerned about that metal being pulled up by the magnets. But then once they take care of the patient, they initiate a standard protocol for image acquisition. So it's not -- they're not -- it's very difficult in ultrasound because if this room was perfectly dark and I was trying to look for a button on your shirt, I'd have to sit there and look and scan and I'd have to know in what direction I was looking at. But then I'd have to understand and interpret. Well, there are now -- we were developing -- we are developing AI models and have received a major FDA approval recently. But we've opened up our ecosystem, and we now have 30 development partners who are building AI ultrasound into an SDK of ours.

And so when they launch it into the App Store, the customer can buy the keys from that company. They plug it into a Butterfly. Now the AI can tell them how to diagnose a DVT, how to look at a cardiac echo, how to be able to do a pulmonary scan, how to do an aortic scan or a fetal monitoring scan. So the power of this ecosystem is insane. And as these apps come out in more and more applications, what's going to happen to legacy ultrasound, especially on the lower end, is similar to what Apple did to BlackBerry. BlackBerry didn't have an ecosystem, didn't have apps. You were all forced to have a BlackBerry because your enterprise managers at the time told you that this was the safe and lockdown system.

But then over time, when Google came out with Google Maps and you can now navigate on your phone, you're like, "Wow, I got to have one of these iPhones." It does all these things and all these cool apps. Well, we're architected that way. And there's going to be all of these AI apps that come into our ecosystem. And it's very, very, very powerful. And one of the first ones, HeartFocus, a cardiac echo app came out about 6 months. And at some point, as these apps come out, this whole thing is just going to explode. And then also, it takes -- it's -- now that this market has been around for point-of-care for about 15 years, and it takes a village. You have to make the device, you have to train on the device.

You have to have these applications. And then you also have to have services. So we do implementation services for our software. We do services for training. And then now we're building a home business to work with at-risk providers to keep their patients in the nursing homes and not in a revolving door to the hospital. And we'll have a lot more to talk about that on our next earnings call.

But ultrasound is having a moment. If you just look at all the different papers, there are -- and this is normalized because we have, from a growth perspective, normal ultrasound has a consistent amount of papers. Point-of-care ultrasound from a relative percentage standpoint is doubling the amount of papers. But there are now a lot of things in transcranial ultrasound, neuromodulation, brain-computer interface, sonoporation, blood-brain barrier opening. These are all things now that are being studied for the first time. And there's a lot of new applications. And that's why we are starting to open up our platform.

And what that does for us is we have just about a $20 billion TAM for core Butterfly. But the moment we start licensing our chip to other markets, it expands our TAM from about $20 billion to about $350 billion. And we think this is just the beginning. And I'll show you in a moment a list of companies or a list of areas that are starting to work on our chip. So we've created a program called Butterfly Embedded, looks a lot similar to an Intel Inside logo. And we are creating this concept that now we have a cloud compute, now we have apps. Now we have a lot of APIs and SDKs to each of our individual systems, and we're making it very easy now to start partnering with third-party companies that not only dramatically increase our TAM, but also increases our own intelligence that as we help solve their problems, we're learning more about our capabilities and where that market needs to go.

So when we look at that overall TAM and that overall opportunity, Point-of-care ultrasound itself was our entire world. That's what we have spent the last 15 years focusing on. And it is an exciting time because in that world of point-of-care ultrasound, our fourth-generation semiconductor will be out next year. And that fourth-generation semiconductor is a profound technological shift in the ability to acquire images. And that will, in my view, if -- probably one of the biggest analogies for our business is digital photography. So if you know the story of Kodak, back in '95, they launched the first digital camera. And every business school studied the fact that in 1997, they shut that program down. And they shut the program down because it was a 1-megapixel image.

So whenever digital equals ultrasound or digital equals ultrasound -- whenever digital equals analog, digital wins. In the beginning when Netflix launched, what was the biggest problem? You had buffering. Your home network wasn't as powerful to run the bandwidth or it was too big a file to push across your home network. And so regardless of it being a cool idea, you still went to Blockbuster because you want to see the thing whatever. But the moment Netflix performed as well as Blockbuster and they were able to compress it and multidimension it, then now it became a benefit.

Well, in 2003, all the other great imaging companies said, "Now, this digital photography is a great idea." And when it got to about 5 to 7 megapixels, digital photography took over film. Well, we are the digital photography part, and we now have seen our processing power double and our ability to do harmonics is now significant through our MEMS capability. So the next version of our probe will be better than any other probe that's out there. And there won't be a need for the other companies to even sell their product because our -- right now, you look at your phones, Apple just launched a phone with a 42-megapixel image. So that's not going to stop. The capabilities of semiconductors aren't going to stop in its development. And we have been rapidly developing, which now allows us to participate in a lot of new markets.

And so we're talking to companies today in patient monitoring. We're talking to pharmaceutical companies today about how ultrasound can do drug activation. There are some bone healing opportunities, endoluminal HIFU, which is high-frequency ultrasound that will allow you to focus a beam to do some type of something interventional or clinical, catheter-based. We have a deal with a robotics company now, neuromodulation and BCI, and that's probably one of the biggest, most explosive ideas that are coming out that they're looking to put our chips in the brain.

And because our chip can move and the beam can move, you can put one chip in a lobe and see that entire lobe of the brain. So -- and by doing so, you can see blood flow, you can see electrical activity. And if you wish, you can actually push thermal to an area of blood ischemia to try to stimulate it. So it's -- what people are now trying to do is, look, we've solved the problem for ourselves. We're building a point-of-care ultrasound business, but we now have a lot of very big shots on goal, and we probably have now 40 to 50 different companies we are in active discussions with to license our chip.

And because our moat is so big, a lot of this has all been inbounds. And because of our IP and because of people's desire to do this, most of this has been inbounds. But we've now, in the first quarter, started to build a team that is now proactively marketing to certain industries because this is not just health care applications as you see around the circle. If you look in that box, there's actually ways to keep batteries fresh by delivering pulsed ultrasound energies. I even had -- on the material stress side, I just met with a company that's putting -- trying to build sensors for drone wings to ensure that with the amount of pressure they're under, they don't snap in air. And you can see with ultrasound material stress changes before it gets -- and it breaks or even cracks.

So there's -- and there's even an application that someone is working on that allows for -- if you have this massive window, you can put our chip on the window and you can touch the window. And when you touch the window, the chip knows exactly where your finger is. So think about that. You can have a spot on your window that's a light switch or a keyboard or you can have a refrigerator with a keyboard and things that you just dynamically touch. Just like those movies where you see what was that thing? I can see it in my mind. Anyways, whatever, but the computers that you can see in there.

So anyways, a lot of really cool things that are happening for the company. And so today, this is kind of our pipeline. So what's active are the ones that we are now signed. So we have a robotic surgery product. We have BCI neuromodule -- 3 BCI companies, diagnostic imaging, a tomography company, which is fascinating. We have an interventional company, a wearable company, augmented reality. And in the pipeline, you can see all the different businesses that we're speaking to. Our chip licensing business, well, first of all, you can see I'm very excited about our point-of-care ultrasound business, and that's going to be a wonderful growth and very exciting for us into the future.

The chip part of the business is going to be a much bigger part of our business. It's going to -- it's today maybe 20% of our revenue after 1 year or after 2 years. And it will very quickly not only grow past that business, but be a major profit driver for us and a major annuity for us because all of these partnerships that we are doing are all incremental. It doesn't require any capital CapEx investment from our side. We just continue to develop our chips, and then we now can license them out. We can sell -- get the revenue for selling them. And also, there's revenue associated to them if they want us to do any individual work.

And so our revenue now is coming from point-of-care ultrasound and what we call embedded. We have our core business. We will now in -- the end of '26, start seeing revenue from this services business in home. I'll talk about that more on our conference call. And then Butterfly Embedded, the way we make money is through a licensing fee, chip purchasing and revenue share or a royalty. And so Butterfly started off as a SPAC, went out really hot and heavy and realized that in health care, my view has always been in health care, it takes 10 years to become an overnight success. But as you look at any of these companies, any of these large tech companies, first 10 years are hard. You're really having to create a lot of foundational work, but then the marketplace gets it and it completely explodes.

Look at any of the bigs right now. I think even SpaceX was private for 20 years before now they're going public. So it takes a long, long time. But we've gotten this thing turned around. We came -- I came here in '23. We reduced the size of the business. We focused on launching our next probe and consuming a lot less cash. And now we should do between $117 million and $121 million this year, dramatically reduce our cash consumption and set ourselves up for real serious growth in the future.

And so last quarter, we grew 25% and you can see our gross margins improving. You can see our consumption of cash reducing. And this is just the beginning. I think we are near that 10-year overnight success mark. I think a lot of people are waking up now to see who Butterfly is in the context of the tech world. A lot of the tech community we're talking to now at the highest levels, guys, we're having conversations with. And I think this company is poised over the next 3 years to do some crazy fun things.

Also, we did a deal with a company called Midjourney. It was a $74 million licensing deal. And that's one of the reasons why we're seeing this type of revenue growth this year. Just stay tuned because a lot of people want to know what they're using our chip for. It's a Gen AI text-to-image company. That's an extraordinarily creative company. And I'm very excited about what they're going to unveil when they're ready. But hopefully, it will be soon. All right, guys. I appreciate your time. Thank you.

[Audio Gap] to Bank of America for giving us the opportunity to present. So my name is Joe DeVivo. I'm the CEO of Butterfly Network, and I look forward to either introducing you or giving you an update to the overall story.

So we are a point-of-care ultrasound business. Point of care ultrasound is bringing ultrasound to where the patients are. So instead of a patient having to go into the hospital and have a radiology exam, doctors are educated to use a device, nurses. So wherever the patient is in the hospital or in the outside, in the home, doctors would carry these devices to be able to manage patents.

And it's a whole new way of reducing the cost of care, democratizing ultrasound because about 2/3 of the world doesn't have access to medical imaging. They don't have that beautiful hospital where these just go in and get the overall image. And so we are bringing -- can I have the clicker, if you don't mind? So we are a semiconductor-based ultrasound device. So that semiconductor acts very much like a TV screen where we are able to program it to deliver all different types of sound dynamically.

So about 20% of patients with serious conditions are first misdiagnosed. About 80% of diagnostic dilemmas can be solved with simple imaging and the initial assessments with imaging results in changed medical diagnosis if they're in the right upfront. That means like at your primary chair doctor, that means when you engage, instead of them taking your symptoms and then making the best decision with that data, giving them more data. It's like we were -- we are the stethoscope salespeople 40 years ago who said, "Hey, doc, it might be good if you to listen to the bowel sounds, listen to the fluid and the lumps, listen to the heart." Now we're giving them a device that's only several thousand dollars to be able to see inside the body when they see the patient. And it's having massive impacts around the world.

And the way it works is supercomputers used to be huge. Now we all carry one in our pocket. Well, ultrasound machines are huge. Now doctors and nurses are going to be able to have a device in their pocket. And it's all made possible by the development of the semiconductor because whether you look at the heart, whether you look at the lungs, MSK, peripheral, vascular, you would typically need a different probe for each one of those applications. And what we have now is the ability to just identify what part of the body you would like to scan, choose that part of the body and then the settings of the chip change to be able to deliver the type of array and frequency necessary for that part of the body.

What we are to classic ultrasound is what digital photography was to film. We are revolutionizing ultrasound through digital, the digital acquisition. Now, ultrasound will do back-end digital processing, but they don't capture the image in a digital form where they can change and dynamically change those settings. So it's a very, very powerful tool, and we're making a lot of progress, so much so that we're actually getting noticed in the world. We're getting noticed. And we are on The Pitt, we were on just recently on Survivor and then Virgin River.

I never watched that show that's my wife show, but she got excited when this particular person had a lower rate quadrant pain and had to diagnose an appendectomy. But what we're making -- Butterfly is making its mark in health care. We have over 150,000 devices sold to date. There are doctors in all different types of walks of life that have them for all their different use cases. And now we're becoming a verb, "Did you Butterfly that patient?" It's really, really strong type of adoption. And now we're working with large institutions about institutionalizing this into the main component of health care.

Now, also, as individual doctors walk around with individual ultrasound devices in order to make this work for hospitals, they have to be able to capture the scan. So we're one of the only companies in -- the only medical companies in the world to ever receive an Apple design award. We are architected in a cloud-based app-based modern architecture with a modern SaaS platform that integrates to all the systems in the hospital.

So we can have -- we have hospitals now who have over 1,000 different devices and all those devices stream data into the stream data into the EMR, stream data into the DICOM, stream data into the revenue cycle management system because in point-of-care ultrasound, hospitals only capture about 35% of their scans. And so that's 65% leakage on revenue and also data compliance as they're they need to have that data to be compliant.

So we've architected ourselves in a very modern way. When I was running other device companies, and I wanted to update my software on a device in a hospital, I have to send my rep in with a USB stick. Now I can push an update just like iPhone to all those devices around the world. We come up with a new concept, we push it. And so our architecture is actually one of our strategic advantages.

And what it allows us is in ultrasound, the hardest part about ultrasound is that it's hard. If you've had an MRI, if anyone has had an MRI or CT, you go into the MRI lab and the tech puts you on the table and tells you not to hit your head on the way in. And then they go outside and initiate an automated protocol. Ultrasound is like an art. You have to actually search the anatomy, and then you have to understand what you found. It's a much harder modality to do.

So it's kind of an enigma. You have this low-cost modality but it's harder to do. So how do you make it easier? AI. There's over 30 companies now in Butterfly Garden developing LLM models to be able to make it easier for caregivers to do an ultrasound scan and at a minimum, make it easier to acquire the image. So the 2 biggest challenges in ultrasound is image acquisition and image interpretation. If you make it easy for a nurse to go into a home and do a cardiac echo, which is the highest volume cardiac image taken, you can send that image to a cardiology clinic, they can make a decision, change their medications for that patient, keep them in the home, keep them in the nursing home, keep them where they're at and stop the revolving door in and out of the hospital because they're not getting the appropriate management of care.

The amount of apps we have now and the amount of -- whether it's cardiac echo or pulmonary or MFM, we just received the first FDA approval of a blind-sweep of any company in the world just a month ago. And FDA was so proud of it they actually tweeted and posted on LinkedIn that this is the way AI should be approved with the appropriate safety, and efficacy data and it's all based upon the apps and how we come into our ecosystem. That will change health care. And just like the iPhone took over BlackBerry, we will take over all of the legacy handhelds when we have all of our applications easy and in a modern way.

We also now are helping work with at-risk providers to be able to show them how to use this in their practice and how to use this in order to be closer to patients. And we are going to be creating a whole new home care business that allows us to manage the logistics in order to move that part of this forward. So that's another we have our devices, we have software, and now we're building a service business. That will be a very material contributor in 2027.

Now there is a moment happening right now. And that moment is ultrasound is being seen as an amazing bridge into the human body by AI, and a lot of different use cases, and this is normalized from a growth rate standpoint. There's about 13,000 to 16,000 studies on normal ultrasound a year, but it's very flat. There's about 2,000 studies on point-of-care ultrasound a year, but it's very flat. But now new applications on how ultrasound can be used in the brain for neuromodulation, how it could be used for deep brain stimulation, where literally we are the only company in the world that has commercialized a digital ultrasound on a chip. And we may be the next bridge between that chip and AI in the human body.

And there's a lot of use cases now. So what we've made a decision is to start a business by licensing our chip to third-party companies who are not competitive with Butterfly in its core space. And that's called our Embedded program because we have solved a foundational problem in overall technology and we're the only one who can put this chip and have this continuously evolving semiconductor-based ultrasound. And what it does for us is our TAM for point-of-care ultrasound is just about $30 billion. But now by partnering with other companies in some very big use cases, our TAM goes from about $30 billion up to about $35 billion because we can participate in all these different segments of the market.

So we've launched a program called Butterfly Embedded. For those looking at the slides a lot like Intel Inside, as you see in the bottom right, and that's exactly what it is, is that we have an ultrasound processor that can help third parties accelerate their ultrasound aspirations by using our tech platform. From 1993 to 2005, NVIDIA was a video game company. They made processes for video games. What they did. They had a GPU, they had POCUS because they had to have this multidimensional activity and then they had to have someone else have this multidimensional activity interact, and they constantly evolve that.

And it wasn't until 2005 when they launched something called CUDA. CUDA was their developer software, and that allowed third parties for the first time to go into the chip and change the settings for their particular applications. That's when the world started to change from them. And if you listen -- I listen to The Joe Rogan Podcast that had Jensen on for 3 hours. I listened to him speak to Stanford students, where he said, "We had no idea what our next chip would be used for. We just kept pushing the boundaries of technology."

Well, we're taking that same mindset. We're a point-of-care ultrasound company primarily. But we now have opened our software up to third-party developers, and they are now finding all different types of new use cases for the technology, and that's turning into pretty potent revenue stream. So a company that was purely point-of-care ultrasound POCUS now has shots on goals for all these other markets, and we are actively talking to companies in each of these markets.

So we are developing point-of-care ultrasound. We have a sales force, we have a marketing team, we have the cost of building that market but we can now participate in all these other markets without having to develop them but by partnering with someone who's there who needs our core technology. We have 600 patents in this space. That's cost us $300 million to make our first chip. We are now selling our third-generation ship. Our fourth generation is at TSMC, who is our partner, and we are developing our fifth generation ship.

And here's our pipeline. We've never shown this before. but we have 4 companies that are working with us that have been publicly disclosed. We have another 5 companies that don't want to be unveiled yet, but these are real early players in their space and those are our partners. And then we have another 40 companies that we are actively talking to license our chip.

So the ability to build our point-of-care ultrasound market and grow and see the dream of every doctor and every nurse having their own device is what our core mission is, but being able to now leverage the technology for all different parts of health care to accelerate the benefit of human care, and then also to have our existing shareholders benefit from all that innovation is why I think the next several years of Butterfly are going to be absolutely insane.

When these applications start going public when these companies start bringing them out and also in point-of-care properly adopts and inflects, this is going to be one of the most exciting growth stories you guys are going to watch. You just remember, that you're here today because this is a real exciting time for us.

The way our revenue as our company comes in, is both POCUS and Embedded. We are now breaking out the Embedded revenue and we sell hardware, services and software in the POCUS segment. And then on the Embedded side or the chip side, we have a licensing fee, we sell chips and we do a revenue share, depending upon different models that they want exclusivity, et cetera. And that within the most recent quarter, it was about -- or in this year, it's about 20%, 25% of our revenue now, and I think it's going to be a much bigger part.

We're in the midst of a big turnaround. If anyone has seen Butterfly, we used to be a company with declining revenues, lower gross margins, burning a lot of cash. Over the last 3 years, we're now doubling our revenue. We're in the high 60s in gross margin. We have a strong balance sheet. We're using less and less cash. We'll get to breakeven on a current cash and we're going to be generating a tremendous amount of value. And just last quarter, we grew 25%, improved our gross margins by 600 basis points, reduced our loss and are really focusing on being very good shepherds of our cash.

So with that, I just want to thank you for your attention and, again, for Bank of America for giving us this opportunity. So thank you very much.

Hello, everyone, and thank you for joining the Butterfly Network's Inc. Q1 2026 Earnings Call. My name is Gabrielle, and I will be coordinating your call today. [Operator Instructions].

I will now hand over to your host, John Doherty. Please go ahead.

Good morning, and thanks to all of you for joining our call today. Earlier, Butterfly released financial results for the first quarter ended March 31, 2026. We also provided a business update. The release, which includes a reconciliation of management's use of non-GAAP financial measures compared to the most applicable GAAP measures is currently available on the Investors section of the company's website at ir.butterflynetwork.com. I, John Doherty, Chief Financial Officer of Butterfly, along with Joseph DeVivo, Butterfly's Chairman and Chief Executive Officer, will host the call this morning.

During today's call, we will be making certain forward-looking statements. These statements may include, among other things, expectations with respect to financial results, future performance, development and commercialization of products and services, potential regulatory approvals, revenue attributable to embedded partnerships through revenue share, ship purchases or otherwise and the size and potential growth of current or future markets for our products and services.

These forward-looking statements are based on current information, assumptions and expectations that are subject to change and involve a number of known and unknown risks, uncertainties and other factors that may cause actual results to differ materially from those contained in the forward-looking statements. These and other risks are described in our filings made with the Securities and Exchange Commission. You are cautioned not to place undue reliance on these forward-looking statements, and the company disclaims any obligation to update such statements. As a reminder, this call is being webcast live and recorded. To access the webcast, please visit the Events section of our investor website. A replay of the event will also be available on this page following the call.

I would now like to turn the call over to Joe.

Thanks, John. So before I begin, I want to take a moment to congratulate our Board member, Dr. Erica Schwartz on her nomination to serve as this country's CDC Director. The entire Butterfly family is proud of her and we remain grateful to have her as a valued member of our Board. We've been so fortunate to benefit from her insight, leadership and friendship for the past 5 years. She has certainly contributed to where Butterfly is today. So with that, I'm pleased to welcome you to our first quarter call this year and to share that we've been making real progress across every facet of our business. Our plan to become the leader in point-of-care ultrasound is firmly established. And the market is moving towards a system-wide understanding of the power Butterfly can bring.

Butterfly opened 2026 with another strong financial performance, 25% revenue growth, 69% gross margin and the lowest first quarter net loss since going public, all while maintaining a very strong balance sheet to support future growth. This performance shows our momentum regardless of the environmental noise. And it's an indicator of what's ahead as we continue to make meaningful progress across every phase of our business and I'm excited to walk you through our updates.

Butterfly is in the middle of the digital transformation of imaging, driven by the power of semiconductors and ultrasound. We are beginning to experience the convergence of massive computational power, artificial intelligence and the incredibly customizable digital visualization, which can disrupt not only imaging and health care, but far beyond. We are living through a profound moment of technological acceleration, one that health care has been waiting for, the merge of intelligence with the human body through smart ultrasound form factors that can live with every person everywhere.

It starts with our core POCUS business, empowering every doctor and nurse with a powerful digital imaging device, connected to an ecosystem of AI applications, education and data to deliver immediate clinical care wherever the patient is in the world. We are excited to be the first company ever to earn FDA clearance of a blind sweep AI tool, which in 2 minutes can determine the age of a fetus. You may have seen this circulating on either LinkedIn or X. But we were very proud that the FDA itself highlighted our clearance publicly on social media as a meaningful demonstration of their increased focus on AI-enabled technologies and how to do it correctly in partnership with them. I have reposted it if you want to go to my account and find it.

The GA tool is a true reflection of our mission. Each year, millions of women around the globe enter pregnancy without a reliable way to determine gestational age. In global, rural or emergency settings, patients often enter care not knowing key dates due to lack of prior prenatal care or are unable to verbalize them due to their condition. Without accurate gestational age, clinical decision-making becomes more complex as each stage of pregnancy requires different treatment pathways.

Now, clinicians can determine gestational age quickly and with a high degree of accuracy using our GA tool that's built on deep learning models from Dr. Jeffrey Stringer's team at the University of North Carolina and trained on over 21 million images. Butterfly is now in the process of launching this tool to those who need it most in the U.S. and many global health settings that rely on FDA regulatory pathways.

Moments like this demonstrate the strength of Butterfly's architecture. With a single cloud-based software update similar to how a smartphone is updated, we can push this capability to existing users anywhere cleared in the world. The GA tool is expected to unlock new relationships with ministries of health across developing countries, moving beyond the idea of democratizing health care to actually delivering on it. AI applications that give local caregivers access to advanced diagnostics will be the most powerful accelerator of handheld point-of-care ultrasound. Whether developed internally or through our partners in the Butterfly Garden, these applications can make the complex simple and expand access to patients in highly scalable and cost-effective ways.

This quarter, 2 new partners joined to develop in the Butterfly Garden, bringing the portfolio total to 30. Four of our existing partners have received FDA clearance for versions of their software. Most recently, Deepecho received breakthrough designation from the FDA and expects clearance of a robust prenatal AI application potentially by the end of this year.

We anticipate iCardio going live with an FDA-cleared model this year, beginning of next year as well. And we are very pleased with the progress HeartFocus is making following its launch of a clinical app late last year. We believe all of these applications, combined with Butterfly's internally developed capabilities will become the largest, most powerful AI-powered ultrasound library in the industry. Over the next 3 years, we expect this to drive a meaningful inflection in probe utilization and overall system adoption.

So continuing with POCUS, our Compass AI launched last quarter is showing early success. Our pipeline of software deals measured in total contract value is up sharply from the first quarter of 2025, serving as an early indicator of meaningful enterprise adoption in 2026. We have already closed our first enterprise deal this quarter and are seeing strong momentum, including a 7-figure software TCV deal with a customer committing to the long-term Compass AI road map.

In addition, medical schools are accelerating one-to-one adoption with nearly 1,000 probes sold so far this year across 6 institutions and more programs coming online. These efforts are supported by continued advancement of our already robust enterprise security posture with recent milestones, including HITRUST r2 certification and entry into the FedRAMP marketplace as in process with VA sponsorship, further enhancing our ability to support large health systems and expand into federal opportunities.

Over the next few quarters, we expect to further the global scale of our POCUS business continuing to open multiple new markets across the Americas and Asia, including some of the largest and fastest-growing POCUS markets like Brazil. At the same time, we've been working to expand iQ3 availability to over a dozen additional countries already in our regulatory pipeline.

So moving to Home. Now officially named Butterfly Home and Community Care, we are making important progress. We expect now our first commercial agreement to be signed in the first half of this year and begin training nurses to support patient scanning across our full state in the third quarter. We believe this model can expand significantly across the United States in 2027 and represents a foundational shift in how care is delivered. It enables at-risk providers to take control of patient outcomes with AI-powered diagnostics, reducing unnecessary hospital visits and improving system efficiency. I'm very proud of our team for maintaining their focus and skilled execution in building this business.

So lastly, I'd like to briefly update you on our efforts regarding the RoHS Lead Exemption for handheld ultrasound. We filed an application to revoke this exemption last year, which currently allows the continued use and disposal of toxic materials in lead-based piezoelectric crystal handheld ultrasound in the EU. The commission engaged a third party named OKYO to review the exemption request and make a recommendation. And through this process, they confirmed CMA to be a technically viable alternative, meaning they have met or exceeded the capability of incumbent technologies. OKYO ultimately recommended reapproval of the exemption, but only for 2 years. This is important because exemptions can typically span anywhere from 2 to 7 years. And therefore, OKYO opted to recommend the lowest possible term for an exemption following its review of our request.

We believe that during these 2 years, the EU will gain the understanding they need to revoke the lead exemption for handhelds, including greater awareness, viable alternatives to piezoelectric crystals and handheld ultrasound devices and clarity that replacing piezo handhelds will not present an impediment to health care delivery, but rather introduce a more digital and more effective delivery of medicine. We understand the complexity of disrupting established industry practices, but believe the EU has gained significant insights into the existing of a viable alternative and meaningful progress is being made. So I'll save our embedded update after John's comments.

I would now like to turn it over to John. John?

Thanks, Joe. Butterfly continued its solid and focused execution in the first quarter of 2026 with an increase in revenue driven by double-digit growth in both our core and embedded businesses, an increase in gross margin and continued improvement in operating performance with further reallocation of resources towards higher ROI opportunities and markets.

Let me touch on a few highlights in the quarter that demonstrate this, including revenue attainment that was above consensus, continued gross margin improvement, adjusted EBITDA that was above consensus and our guidance range with improvement driven by our revenue performance, higher gross margin and continued financial discipline. Growth in probe sales of 5% and an increase in the mix of iQ3 devices driving an 11% improvement in ASP year-over-year. The expansion of the Butterfly Garden platform as well as the inclusion of the GA tool in our latest software update, execution of a large 7-figure TCV Compass AI contract with a growing pipeline in this area that Joe mentioned upfront and the expected commercial launch in an initial state with a major U.S. direct care provider to support in-person, virtual and in-home health care services.

With that, let me move on to our results. We started the year strong with revenue of $26.5 million, an increase of 25% year-over-year. Our growth was primarily driven by strength in our U.S. health, international and vet channels, a higher mix of sales of the iQ3 year-over-year, driving a higher ASP and from Butterfly Embedded. Breaking things down between the U.S. and international channels, during the first quarter, U.S. revenue was $21.4 million, which was 25% higher year-over-year, driven by revenue from Embedded as well as solid demand in the core business with unit sales up 5% in what is typically a slower quarter due to seasonality. Total international revenue increased by 23% year-over-year to $5.2 million in the quarter.

While sales of the iQ3 in the quarter were up 39% year-over-year, sales of the lower-priced iQ+ were down 43%, driving the increase in ASP. With the increasing contribution from Embedded today and going forward, I'm going to provide you with a revenue split between our core business and Butterfly Embedded. We have also included this split in our 10-Q.

The core business includes probe sales and related software, Compass AI, other services and in the future Home. Embedded revenue currently includes onetime NRE payments, annual license fees, revenue from SOW-driven development work and chip sales to Embedded partners. With that, core revenue for the first quarter was $20.8 million, an increase of 10% versus the first quarter of 2025. This increase was driven by growth in volume and price with a higher mix of iQ3 sales, which strength in U.S. health, international, e-com and vet. Butterfly Embedded revenue was $5.7 million, an increase of 147% versus the first quarter of 2025. This increase was primarily driven by the Midjourney partnership.

Moving on to gross profit. Gross profit was $18.3 million in the first quarter, a 37% increase as compared to the prior year gross profit of $13.4 million. Gross profit margin percentage increased to 69% from 63% in the prior year period, an increase of 9%. Gross margin percentage was positively impacted by the higher-margin Butterfly Embedded revenue and the higher mix of iQ3 and its higher selling price, along with lower software amortization.

Moving to EBITDA and cash. For the first quarter of 2026, adjusted EBITDA loss was $6.1 million compared with a loss of $9.1 million for the same period in 2025, an improvement of 32%. The improvement in adjusted EBITDA loss in the first quarter was driven by contribution from higher-margin revenue and continued financial discipline. Our cash and cash equivalent balance, excluding restricted cash at the end of the first quarter was $138 million and the use of cash in the quarter was $12.5 million. This compares to a use of cash of $14.7 million in the prior year quarter, an improvement of $2.2 million or 15%.

These results continue to demonstrate that we are very well positioned as we move forward to continue to invest in the business in areas where we see significant opportunities for additional growth and disruption, including expanding our POCUS business and penetration of Compass AI as a core operating system for health systems, empowering third-party development through Butterfly Garden to accelerate adoption of our platform, expanding our Home and Community Care business into its commercial phase. Enabling a new wave of ultrasound-on-chip technologies through Butterfly Embedded and continuing AI and semiconductor innovation with the development of our fourth-generation chip.

Before turning to guidance, I want to update you on the global macroeconomic environment relative to Butterfly. We continue to monitor the war in the Middle East and the ripple effects on the global economy as well as tariffs in certain markets. While there are some impacts to our business, they have been minor. And we continue to manage through it and make the appropriate adjustments. Our first quarter 2026 results are indicative of this. And our second quarter and full year 2026 guidance include any expected impacts.

I would now like to turn to our outlook for the second quarter of 2026 and for the calendar year ending December 31, 2026. In the second quarter, we expect revenue in the range of $27 million to $31 million or a year-over-year increase of 24% at the midpoint. We expect an adjusted EBITDA loss in the range of $6 million to $8 million. For the full year 2026, we are reaffirming our guidance for both revenue and adjusted EBITDA. We expect revenue to be between $117 million and $121 million, an increase of approximately 20% to 24% over 2025. We expect our adjusted EBITDA loss to be between $21 million and $25 million. As I mentioned previously, our guidance for adjusted EBITDA in the second quarter and full year includes increased investment in key areas to support continued innovation and revenue growth in our core business and our emerging Embedded business for 2026 and beyond.

In summary, we had a strong start to 2026. We came in above the midpoint of our revenue guidance and we beat our adjusted EBITDA guidance. We believe we are very well positioned for the balance of 2026. As our 2026 full year guidance indicates, we look forward to continued growth this year and beyond. Our overall outlook on the business is positive with the first quarter reinforcing our view. We continue to be focused on gaining share in 2026 through deeper penetration of existing customers, new customers and applications. We also continue to be focused on enhancing our existing and developing new partnerships, leveraging our ultrasound on-chip platform.

We continue to be excited about the potential of Butterfly Embedded and our licensing and related revenue opportunities. The business is getting stronger, with core focus generating new revenue opportunities in addition to probe sales through Compass, Garden and Home and the potential of Butterfly Embedded. This is happening while we continue our intense focus on driving operating efficiency across the business and return on investment. I continue to be excited about what is ahead for the company in 2026 and beyond.

Now, let me hand it back to Joe for some closing comments.

Thanks, John. Ultrasound has many use cases, which are growing rapidly in both scientific and clinical importance. As I mentioned last quarter, if you scan publications in nature over the past year, you'll find a body of literature growing around functional ultrasound. These applications are ranging from neuro imaging and brain computer interfaces, biosensing and targeted drug delivery to wearable monitoring and autonomous robotic procedures, just to name a few. Who was once a diagnostic modality is aspiring to become something far more powerful. It's becoming a dynamic interface with biology itself.

As I mentioned at the start, what is driving this shift now is the convergence of massive computational power, artificial intelligence and a new class of digital imaging systems. For the first time, imaging is no longer static. It's becoming adaptive. It's becoming intelligent. It's beginning to learn. There is only one company in the world that has commercialized ultrasound at scale through a fully digital semiconductor-based system that is infinitely programmable. That's Butterfly. Our traditional piezoelectric crystal, well, they're fixed. They transmit and receive in a fixed way. But our digital ultrasound is fundamentally different. It can transmit and receive signal, interpret what it sees through powerful compute and AI and immediately adjust how it transmits and receives again.

It creates a closed-loop system where sensing, thinking and acting are happening continuously in real time. This is the beginning of imaging systems that do not just observe the human body, but will integrate into it. Over time, these systems will move from episodic use into continuous presence, embedded into workflows, into devices and eventually into everyday life. This is where imaging, computation and biology converge into something entirely new.

Butterfly is at the center of this convergence, which is becoming increasingly evident in the brain computer interface space. Our chip is uniquely well suited for this application given their programmable architecture, ability to both transmit and sense with precision and tight integration with software compute and AI. This is exactly where our Apollo platform is headed. By extending this foundation, we are building a platform that combines the ability to image and stimulate through software and enable new classes of applications beyond traditional handheld imaging.

We are investing strategically in the talent, the partnerships and the platform to lead in this new category through Butterfly Embedded, which we introduced last quarter as a rebrand of Optiv. Embedded is being led by Dr. David Horsley, a proven innovator in MEMS and semiconductor systems. And we are furthering our engineering and commercial capabilities required to engage with the most important technology platforms in the world.

Establishing the Embedded business with intentions of building a team in San Francisco is a critical step in extending our platform into the broader technology ecosystem. The early traction in Embedded is very promising. We signed our ninth partner this month and expect to add at least 2 more mid-year. And these are not just incremental relationships. They represent the early formation of a new market where ultrasound becomes a native capability inside other systems and we are just getting started.

Speaking of transformative platforms, we are very excited about the progress we are making with Midjourney and look forward to them unveiling their solution when they are ready. It is clear to us today that the Apollo chip will be a major catalyst of revenue growth for Butterfly. Not only will it usher in the next wave of AI in image processing and enhancement as well as on-device AI compute for POCUS.

Our partners want the processing speed it will provide given the massive amount of data they intend to generate. As I mentioned, not only will Apollo have the more powerful MEMS elements that enable harmonic imaging, it will be able to process 20x the data of the chip that we will launch next year to 5.1. This new architecture is being designed into very sophisticated future systems that stream real-time ultrasound data straight into the GPU memory, providing the ability to process a massive amount of image data far exceeding today's standards. We also have identified versions of the Apollo architecture, which can be modified to meet some of the largest future use cases known today.

When you step back and look at Butterfly today, you see 3 engines of growth that are beginning to reinforce each other. Point-of-care ultrasound is scaling across health systems globally. Home and Community Care is expected to extend that capability directly into the patient environment, changing how care is delivered and paid for. And Embedded extends Butterfly beyond stand-alone devices by integrating ultrasound as an AI-enabled sensing platform into entirely new applications across health care and adjacent markets.

Together, these are not separate businesses. They are all playing a key role in leveraging our unique ultrasound-on-chip ecosystem. Over the next 3 years, we are not just expecting growth. We are entering a period of transformation. As these systems mature and converge, they will unlock new behaviors, new standards of care in entirely new markets that do not exist today. What has historically been episodic, expensive and limited will become continuous, intelligent and broadly accessible. We believe this shift will fundamentally change how the world interacts with imaging with health care and ultimately with the human body itself.

Over the next 3 years, we expect this convergence to materially expand our addressable market and drive a step change in utilization, which we believe will be reflected directly in revenue growth and operating leverage. This is the moment where decades of scientific progress meet the scale of modern compute and the power of artificial intelligence. And when that happens, change does not move linearly. It moves exponentially. We believe Butterfly can become the default imaging and sensing layer for the human body. Butterfly is positioned at the center of this shift. Importantly, we are building this with discipline, maintaining the strength of our balance sheet. So our investors can fully participate in the value creation ahead.

So this call marks my third anniversary with Butterfly. The last 3 years, we were focused on building a framework for success and the foundation to execute. The next 3 years are about running fast, scaling revenue into the future and being the company to enjoy the full potential of what we've built. We are in a new era where imaging becomes intelligent, adaptive and ever present, where it moves closer to the patient becomes part of everyday care and ultimately becomes part of everyday life. We are incredibly excited about what lies ahead. This is only the beginning.

With that, operator, please open it up for questions.

[Operator Instructions] Our first question is from Chase Knickerbocker from Craig-Hallum Capital.

This is Jake Soucheray on for Chase. It seems like the one-to-one partnerships are going well. Can you speak to any progress made with the med schools so far in the second quarter and how that compares to prior years?

Well, the second quarter is happening right now. Usually, the second quarter is our biggest medical school quarter given that their fiscal year ends on June 30. And so yes, I mean, this is -- we had a good first quarter with medical schools. We'll have a very good quarter in the second quarter with medical schools. And the key for us is continuing the momentum to convince them that every student when they come in, should have their own probe.

There are 100,000 medical students in the United States, 25,000 enroll every year. And that is an opportunity if it extended to every school of doubling the amount of probes we sell every year. So we are -- we have very good relationships. We just had weeks ago, the e-com meeting with a lot of people participating in our Lunch & Learn and or AACOM, meeting. And the relationships, the momentum is all growing. So I'd expect our medical school activity to be very positive in the second quarter.

And then I'd like to hear some more color on the Compass AI enterprise deal. How did that process go? Can we expect more of these partnerships materially contributing to revenue in the rest of the year?

Yes. As I said in the prepared remarks, we -- our pipeline has like clicked up. It's -- I think the word in the script was it was up a sharp increase. We're seeing a lot of activity, a lot of receptivity in the innovation that we brought to the platform plus just the general market building. So our pipeline is up pretty big. And if the pipeline is up pretty big, then that's an indicator, the enterprise sales will follow. So I'd expect this to be a very good enterprise year. I'd expect this to be a very good Compass AI year. I think we hit the mark with making this easier for clinicians, faster to document. There's a lot of other tools that we're building. So very, very good start.

Our next question is from Josh Jennings from TD Cowen.

Congratulations on the strong start to the year. Exciting times. I appreciate the download and Joe, you're sharing your enthusiasm on all the progress in the future. I wanted to ask about Butterfly Embedded as 9 companies in the portfolio. It sounds like there may be other partnerships on tap. I don't know if there's any way to help us think about the pipeline and any qualitative or quantitative help thinking about it just in terms of -- I mean, should we be thinking that there are more Midjourney type deals on tap, a mix? Any more color on that Embedded partnership portfolio and pipeline would be great to hear.

That's a great question, Josh. It's good to hear from you. Thank you. So I guess the way I'd put it is every time we add a partner, we're adding the opportunity of developing an entirely new market segment, an entirely new market segment where we are embedding our technology and their technology and they have a vision to be able to grow and grow and invest. And so each new partner is an opportunity to create significant leverage for Butterfly. And it really depends upon where the stage of the idea is and where the stage of the company is.

These companies are developing these new technologies. And they also want to keep it secret. They're not looking to give breadcrumbs to their competitors. And they're looking to do their development. So it is a very delicate balance for us where we're a public company. We need to be able to communicate our progress. And we try to do the best we can to frame the progress to give you the understanding of what's happening without getting any trust from our partners.

I mean even the Midjourney deal, as we discussed, was never announced, just a press release. We had to do an 8-K because it was a material deal. But they want to be able to announce and communicate when they're ready. What I will say, because I know what you're getting at is you're asking if any one of these deals that we are working on could potentially have the financial impact in the near term that a Midjourney deal could have? And I would say yes.

We are working on opportunities that could present similar type of near term. I think each one of these are different. Each of them are different companies, different markets and different phases. But sometimes a company comes in and they don't want to invest this type of work unless they have the opportunity to be exclusive in their area. And the exclusivity is what they're paying for. And I would say, yes, there are conversations that we are having that could replicate a Midjourney type of transaction.

Appreciate that. And maybe a follow-up just on the update on the Home and Community Care channel, commercial agreement first half, initial statewide deployment in 3Q. I was hoping to better understand your vision of -- or how you envision this playing out? I mean, is the initial statewide deployment of a commercial pilot and then build from there? Or how do you -- how should we be thinking about the contributions from that Home and Community Care channel state by state over the next 6, 12, 24 months?

No problem. It's a very good question again. So as we had communicated in the past, we had conducted a pilot and it was, let's call it, more of a feasibility pilot where our partner wanted to see if this would work. And as we communicated in the past, I think there's about 100 patients that went through the protocol. And we saw a meaningful reduction of admissions and readmissions into the hospital. And so we were able to, in that feasibility work show that this conceptually could be very valuable. And so now we're on to the next phase, which is our first commercial agreement. And if we get everything done, which we expect will now with the activity in place, be done by the first half of the year. That will then have us now operating with the teams commercially.

And instead of just into a controlled pilot, it would be into a real-world setting where there would be education, there would be daily interaction through our platform. And then there would be the ability to measure outcomes. And if that commercial -- the initial commercial stage continues to show the type of economic benefit that the initial pilot stage and it's only logical to believe that we would expand through more states. And we would expand throughout the entire country. And so that's what our goal is for 2027.

So for 2026, we want to get the first state up and running. We want to show this is operational. We want to show this is incremental. We want to show this is profitable. We want to show it's better for patients. And the overlying narrative here by teaching nurses who are at the bedside, caring for these patients every day, how to use AI to help do a scan that they were not classically trained to do, but to do a scan that now allows for a specialist to give expert opinion to that patient where they are is a completely disruptive new model.

The alternative is just simply waiting until they get sick enough where you have to send an ambulance to the skilled nursing facility, transport them to the hospital. We will care them up to radiology, do a cardiac echo, get them into the hospital, give them diuretics, monitor them and then send them back home. So we send them back to the facility via that same ambulance that's up in there. That is not an inexpensive. That is not efficient. And that is not welcoming to our health care system or what is best patient care. Being able to have that expert cardiology cardiologist opinion to happen real time, to have the data extracted real time to allow a lower cost health care professional to capture that image and be able to get that opinion within 24 hours and change the care for that patient and dial them in better, faster is a significant reduction in the overall cost of health care.

And it's just the beginning because we're talking now about managing congestive heart failure patients. But what about patients with bladder infections or we can do 3D bladder scans, wake in and there. They can do one after the other while they're with the patient. With all these new AI tools we're talking about, there's a flywheel here. There could be -- now, for example, they can test for a deep vein thrombosis. They can test for a pleural effusion. There's all different types of things.

Now this is not just about creating a single revenue line in these facilities. It's about accelerating the evolution of handheld ultrasound at the patient bedside by not just selling a probe, but by teaching them how to perform the application and being there with them and creating the linkages between them and expert clinical care and then being there to also measure those outcomes. This is a very, very big deal because what this will do is help accelerate the flywheel of adoption and in its core. So I'd expect to see some revenue contribution in the second half of 2026. And then I'd expect us to be successful. And I expect us to go more and more states and to start to build revenue, very positive growth in 2027.

Our next question is from Andrew Brackmann from William Blair.

Maybe just a follow-up to Josh's first question. I just want to follow up to Josh's first question. So can you maybe just sort of talk to us about the life cycle for the Embedded partnerships? Just how do these typically start? When is that aha moment? And I guess from your vantage point, what's your visibility look like throughout all these phases, visibility into sort of their ultimate use case and how big these could potentially be for you longer term?

I'll do the best I can for that one. So right now, up until 3 months ago, we have been simply getting inbounds. We have people who are looking for solutions. They understand that we have a chip. They bumped into our intellectual property. They call us and they asked us consider working with them. And I think I mentioned before, we have over 40 companies that we are actively speaking to. And they're in a bunch of different markets. It's in pharma. It's in DCI and neuromodulation. It's in robotics, in surgical robotics. It's in HIFU. It's in organ preservation. It's in these different markets where they're already working with some level of ultrasound, but kind of have a closed-loop system.

They want a system that can learn. And also wearables and monitoring in a lot of different clinical variations are being investigated. Some of these are start-up companies who just got their funding. Some of them are larger companies who are looking to add this into their product portfolio. And so each of them have their different time lines. And so they come to us and then they want to understand, integrate with computational systems, how fast can they get the data, how fast can they process the data, where could they process the AI because a lot of this has to do with pulling data out of the body, processing that data and then making clinical decisions or even using the ultrasound to continue to get data to continue to refine it and then even provide therapy through the ultrasound or additional diagnostics.

This is very, very complicated systems. And it's unbelievably exciting. It's all being based upon this new mass computation and also the intelligence of AI. Classic ultrasound systems cannot modify and modulate in real time. If you have a lens that is cut to see near on a camera, say, no matter what you do with your digital, no matter what you do with your energy. When you push it through that glass lens, it's going to only see near. With Butterfly, you can look, you can learn and then it's almost like you're asking another question through ultrasound. And you can change how you ask that question. You can change how you modulate it. You can change the array. You can toggle from 2D to 3D. You can scan, you can slice.

And so what we do is when partners come to us and say, well, we're really interested. The first thing they do is they give us a set of parameters. They say, look, we're trying to get this frequency, this array or in this location. Can you do it? And we say, depending upon those parameters, we will answer that question. And if the question is favorable, we then go into a laboratory setting. And we show them how we can tune the ultrasound. We show them how we can get close to their use case or get to their use case with tuning the chip.

It's kind of like just putting a program on a TV screen. It just reprogram it and then it goes. And then when we prove to them that we're either there or close to there, then that determines the next phase, which is what does it take to get the product and the software to be exactly the way they want it and what would the integration be in their systems. And then we map out a road map. And it's -- here's the work that they would do, here's the work that we would do.

That turns into depending upon the stage of the business, a development agreement. That development agreement would mean that they pay us a licensing fee for our software, just like with NVIDIA, you're paying for a license for their CUDA. And then they'll buy chips and they'll do labs and they'll do development. And if they ask us to help them refine certain things, then they'll pay us engineering fees to be able to help them do that. And then there's some point in time where they realize, well, okay, I have a line of sight to what this means to us commercially. And they will either who like with NVIDIA will just purchase their product and use it or they will be someone to say, well, gosh, if I can be the only one in the world to use it for this use case, would you be willing to license it specifically? And that's where those conversations go.

So these are multiple steps. But what this means is Butterfly shareholders are now tapping into entirely new markets that are well beyond point-of-care ultrasound with opportunities to have very vibrant and profitable recurring revenue streams without having to do the market development that we're doing in point-of-care ultrasound, without having to hire the sales force and supporting a public company and back-office costs. It's simply we get license fees. And we are able to sell our chips. And we now have an option to grow revenue in entirely new markets that would take us a lot more capital than we've had to deploy so far in order to build. So it is an exponential flywheel.

And so we've been developing. And fortunately, for Butterfly, from day 1, we've built our architecture to support this. We are cloud-based. We are app-based, iOS, Android. We integrate with SaaS. We have our hardware. And so, a lot of these components turn into the application that our partners are looking for. So they're not just licensing our chip, but some of them want to use our cloud. Some of them want to use our SaaS-based Compass platform for data aggregation. And so these partnerships are very viable, but it does take time to onboard. There is diligence, there is lab work. But -- and so when we say a new partner comes on, that's a big check because that is a wonderful opportunity for us to increase the TAM of Butterfly and to bring more revenue and market opportunity for our investors.

Great. Appreciate all that color. John, maybe one for you just on guidance here. It seems like there's a lot of tailwinds here between the Garden expansion, the Compass AI contract moving into Home Care in the second half of the year. Can you maybe just sort of unpack for us the underlying assumptions? What's in guidance now from a revenue perspective? And what's the potential upside lever?

What's in guidance now is everything that we have a line of sight of, including across our core business with probe sales. Joe talked to Home. We have some of that in there as well in the second half of the year. We talked about Compass AI. So the things that we have a very good line of sight on are in our guidance. Other things that we might be in the process of negotiating, for example, like if you look at that could be potential upside down the road. It's early in the year. As you know, we just printed the first quarter. So ultimately, I'd say our posture is going to be a little bit more on the conservative side. But we reaffirm guidance. We're very confident in where we're going for 2026.

And in addition, if you look at adjusted EBITDA, obviously, we had a beat in the quarter. But ultimately, we are investing in certain areas of the business. So with the launch of Home, it does require some investment to get it off the ground. And ultimately, it will be a positive. It has a very strong ROI for us, but that takes time. We're not in this to just have a really, really solid 2026. We're also in this to continue to grow at the rates we're growing at for the foreseeable future. We're talking about years here, not just quarters.

Our next question is from Ben Haynor from Lake Street Capital Markets.

First off for me, obviously, great to hear the FDA enthusiasm on the gestational age AI tool. Presumably, this derisks kind of the future blind sweep AI like tools from you guys. But I would imagine that it also does so for folks that might be looking to develop something similar for Butterfly Garden or Butterfly Embedded. Any more color you can kind of share there? Has it peaked more interest? What has that done for you guys?

Well, I think what is important, and I think what the FDA is pointing to is there are a lot of people can imagine right now and I don't have access to the data. But I would imagine there are thousands of applications in the FDA for clearances of AI tools. And I would imagine that a very high percentage of those are probably not doing it in the manner of treating it like a device or treating it like something with consequence because people at times just think software is software.

But when you have to -- when you're going to inform a patient and that information is going to change the clinical pathway, right. And so we have taken a longer path with the gestational AI tool. We have done, is we've done a lot of clinical work. We've done a lot of work with the University of North Carolina and Jeff Stringer, we've done a lot of work making sure that the -- because remember, this blind sweep, there's no image that anyone is looking at. Blind sweep the pregnant abdomen and it spits out an age. And so it's got to be right. And you're not doing it based upon looking at an image. So I think what the FDA saw was that here's a company who took this seriously, understood the consequence. The incredible upside value proposition and did the work to validate the AI just like you would do the work to evaluate any type of device that was going to intervene into the body.

And so I'm proud of our regulatory and quality teams for holding such a high standard. I'm proud of our management team for not being short-minded. A lot of times, people get in and they argue with FDA. They try to convince them that it shouldn't be this way, it shouldn't be that way. And sometimes FDA is wrong. They're not perfect. And sometimes they need persuading or educating and whatnot. So I'm certainly on that front.

But after my 30 years of putting products through FDA, I have -- I think the world is better because we have FDA. FDA keeps us safe. I think FDA keeps us on our toes. And there's always someone who's wrong or someone who might overdo it or underdo it. But the process, I think, leads the world in making sure that this technology works. And when you get through that gauntlet, then you now have -- you have a moat and you have -- because you've earned that moat. And we have that over every ultrasound company in the world. So we are now actively working with partners and ministries of health that have been waiting for this for years.

If you've heard of some of the visionary statements from the Gates Foundation, they've always believed that AI and ultrasound is probably the biggest catalyst to improvement of overall care for these developing countries because of the power of these diagnostics in these remote areas. And I think this is a tool that is going to now translate into revenue by there being large deployments of probes and systems in all these different places. We accomplished something 1.5 years ago, 2 years ago that was extraordinary.

We trained 1,000 midwives in Africa who are managing over 80,000 women who are having childbirth and they're doing, I think, about 80,000 scans a month to make sure that they identify the fetal position of a baby because 20,000 women a year die in childbirth just because they were not able to get the baby out, which is horrible. Like you can't even imagine that. But with the Gates Foundation, we've trained those 1,000 midwives. And they're now improving care every single day. With the architecture that Butterfly has, we can just push this new tool to all thousands of their probes and get them now being able to identify the age earlier.

These are significant flywheels in the acceleration of patient care and that's being done on the architecture of Butterfly. So we're proud and for the FDA to go to their social media and their social media part to specifically call out Butterfly. I think it makes us proud. And it makes us feel like we're doing something right. And that we're a real company who really cares with terrific technology, who has profound respect regulators. And I think they've returned that respect.

FDA communicated is definitely feather in your cap. Overall, just a fantastic update here this morning. The one thing that maybe I missed and I apologize if I did. Any updates on kind of additional form factors? I know historically, you talked about the IQ station and things like that. Anything else on that front? That's it from me.

Yes. So thank you, Ben. We appreciate the questions. So we didn't call it out in this particular call. But as an update, early next year, we will have a new probe. And we will detail the capabilities and the branding and the cost and everything. I think probably closer to the end of this year, beginning of next. So that is online. And that will bring to market not only the -- bring our ultrasound on-chip technology. But it will be the first ultrasound on chip that has harmonics with CMA. So we're thrilled with that innovation. And we are actively working on our station. And that is also in early to mid-2027 time frame. So we didn't chronicle it, but it is very much still in our work. And also, we've talked about wearables for a very long time. And we have a wearable today. But if we went and launch that wearable, I don't know what that market would be for it yet.

What we are doing is building out applications in Home. And then quite frankly, several of the Embedded partners are wearable partners where they not only want the ultrasound chip. But they're developing a wearable form factor around what we've created. And then they are developing applications for clinical application of that wearable.

So I would think that the first wearable that comes to market would be through one of our Embedded partners into some extraordinary use cases that they're developing. So we are -- we do have a lot of resource focused on new technology development and software and AI and hardware. But we are able to accelerate our market opportunity as we partner with other companies with visions and novel ideas for them to be able to create whole new markets. And so those are probably the devices that you'll be seeing from us over the next several years.

We currently have no further questions. So I will hand back to Joe for closing remarks.

So everyone, thank you so much for the interest. I know that the macro environment has a bunch of ups and downs, but I think we're going to be swimming upstream of those. We have a lot of tailwinds. We had tremendous growth last year, tremendous probe growth. We have tremendous new product launches. We are coiling up opportunities for Garden and Embedded. We are coiling up opportunities for more one-to-one medical schools and enterprise. This is the time when we are -- our foots on the accelerator. We're stretching our legs and we're making things happen. And we appreciate your support and the unbelievable energy of all of the employees of Butterfly who are actively changing the world. So thank you all.

Thank you. This concludes today's Butterfly Network, Inc. Q1 2026 Earnings Call. Thank you for joining. You may now disconnect your lines.

Okay. We're going to continue down the morning track, Day 2 of the 46th Annual TD Cowen Healthcare Conference. I'm Josh Jennings from the TD Cowen Medical Devices Research team. And we are excited to have executives from Butterfly Network. Local trip for you guys, but still much appreciated. Thanks for participating in our conference this year and letting us host this fireside chat. We have CEO, Joe DeVivo; and CFO, John Doherty. Gentlemen, thank you, and good to see you.

Good to see you. Thank you so much for the invitation. We're -- not only the conference is a fantastic opportunity for us, we're very appreciative for all your support of Butterfly. And so we wouldn't miss something like this for the world.

Well, we've been intrigued by Butterfly technology initially, the point of care.

Yes. You came on early. You came on early. You paid your dues.

We have paid our dues. The handheld ultrasound market and your technology. But as you've laid out recently, you're not just a point-of-care ultrasound company, and there's been some evolution that we'll dig into. But maybe to start, I've just seen you come into this conference room, I had a little flash back to 2024, that Investor Day event you held and where you and your team laid out kind of the vision and the path. And over the last 2 years, you guys have delivered and executed on multiple fronts, pipeline innovation, this Octave Now Butterfly Embedded initiative. And there's a lot to talk about today, a lot of progress in 2025, and 2026 seems to be shaping up to be an even bigger year. So congrats on all that progress.

And maybe we can start on just Butterfly Embedded because that's a hot topic. You guys announced a deal with -- a partnership with Midjourney last year, big revenue stream. You guys already talked about some of the contributions in 4Q and on a go-forward basis. But maybe any updates? And I know that it may be a little bit limited in terms of what you can share, but any updates on that Midjourney collaboration, maybe to start just on the technology front and the product that's being developed through this collaboration?

Well, thank you for that, and thank you for showing up to the Investor Day meeting, and thank you for listening. Because actually, if anyone is listening, they'll realize we're making amazing progress. And we're very appreciative for the type of support that you've been giving us and other partners. We're on to something really special, guys. And again, for anyone listening, you should just be aware of what's happening. We -- our founder set out to create a technology that had flexibility, that had ubiquity, that had a concept where all the different imaging and ultrasound use cases could be generated within a single device, with a semiconductor that was infinitely programmable, that was low cost, that was durable, that was mobile. And again, we -- a lot of us put our context into big metal med tech and especially the investor base is used to big metal med tech.

And we are not big metal med tech. We are a technology company that is solving a health care problem. But we've learned that we've solved the problem in a way that is probably more foundational than we thought. And the manner in which sound has been delivered over the last 40 years has been through delivering energy through a crystal. And then that crystal would vibrate based upon how it's cut into a certain wavelength based upon the frequency and the energy that's been put through it. And that's kind of like -- it's like saying I've been taking a photograph with this glass lens for the last 40 years.

Our lens now is made out of 9,000 MEMS elements that can be controlled individually. And it was specifically designed to be in a handheld device that allowed for a clinician to say, I want to scan the carotid, I want to scan the heart, I want to scan the MSK. And it would just simply change the settings to allow a clinician to be able to make a clinical determination right then and there. And so that part of the business is growing and the whole mission is strong and what we are accomplishing is impressive. We were on Survivor the other night. I don't know if anyone watches Survivor, but literally 3 nights ago. So for those of you who have...

Not as contestants.

Not as contestants, no. But we -- for -- about 3 nights ago, the opening -- the series premiere for Season 50, one of the contestants had an Achilles injury. And you got to remember that while this is a TV show, this is filmed out on an island in Fiji, right? And so you have this kid trying to wonder if he has a torn Achilles and they pull out a Butterfly right on the show. And they walk through, hey, this is a really potential serious injury, we need to scan this injury, et cetera. And many of you have already seen that we were on the pit on both the Season 12 and Season 13, where they use it in a mass casualty event. It was used in the space station recently. You've seen some news. Inspiration 4 when the SpaceX astronauts were -- it was a full civilian crew with the same -- they used to be on the Board of St. Jude with the St. Jude patient, they had it on board there.

And so we've created this amazing technology that can be anywhere a patient is and can help them substantially. But we've learned that, that architecture or that chip can help other companies to -- and reach their goals, too. We've learned that, that chip being infinitely programmable. So we have the ability -- in a medical environment, but we have the ability to create an unlimited, based upon our parameters, type of sound from a semiconductor. And there are a lot of other companies who would like to use that concept.

And so we're very excited for what that means. Midjourney has come up with a use case that they're very excited with that we cannot unveil until they're ready to unveil. But we're talking to a lot of people on how the interface between technology, AI and the human body can be bridged through the portal of sound. And we're becoming a pretty important part of that. And it's very exciting. And I think it's core to our founder's mission, which was the democratization of imaging, to be able to bring imaging easier, lower cost to where everyone is. I think making sound in that and the interface with the body is going to be an amazing use case as we see cloud AI morphing into physical AI.

And maybe -- you just peaked my interest of cloud AI translating into physical AI. Can you help a lay person like myself, and I'm not a tech guru, but just better understand that dynamic or that -- any more details you can share there on that transformation?

Well, so you have OpenAI, you have Gemini, you have Grok, Claude. In many senses, going to a website and searching for an answer in the cloud is very quickly becoming generic. But now you see what's happened with OpenClaw, where you can create this agent and this agent has all of this incredible capability. Well, what's the first thing everyone wants to do with this agent. They don't want to set it loose on all of our data. They want to buy a separate, I would call it a PC. They're actually buying these Air -- what are they, those little Mac -- iMacs, I think, or iMac minis, I believe they are. And they're creating this whole kind of closed environment and then creating accounts on that piece of hardware for the AI to live.

You're going to -- I would say most AI players today are looking for, down the road, a way for a unique device to interface with a human. And that is kind of the next phase. I mean you've heard of Jony Ive over at Apple has come into OpenAI about a year or 2 ago to create this new device. And you're going to see more and more AI empowered devices interface with a human in order to do a very specific use case. I was listening to that All-in Podcast with Chamath Palihapitiya, and he was specifically talking about the reversal of cloud because now you -- because we put all of our data into the cloud, well, now everyone is going to want to start pulling their data off the cloud and then compartmentalizing it and then running it with these agents that would interface with devices. So it's really, really fascinating what's happening right now. And to the extent that communicating through sound is an important portal, it leaves a massive opportunity for Butterfly to leverage its core technology for these use cases.

I appreciate that. That's helpful. And we are anticipating, I guess, Midjourney's unveiling of the use case and the integration of Butterfly's Ultrasound-on-Chip semiconductor technology. And any time lines you can share just in terms of when that announcement could be made?

Whenever they're ready. We are their partners. It is their business. We are only here to empower them. And when they're at the time that they want to talk about this, man, I can't wait to talk to you, Josh. I can't wait to unpack it for you because it's just so cool, it's so impactful to humanity, and it's this marvelous kind of [ compilation ] of what our capabilities are and what their extraordinary capabilities are and how they kind of amplify each other.

All right. A lot of anticipation building. So we're looking forward to learning more. It's my understanding, and you've shared this publicly, just about you working with Midjourney over the course of 2025, maybe even late 2024 in front of finalizing this collaboration. But in that period, there are other potential partners around the hoop for what was once branded Octave now Butterfly Embedded and just this finalization of the Midjourney partnership may have catalyzed even stronger interest. And maybe just talk about the pipeline. You referenced this on the recent earnings call last week, but within Butterfly Embedded, could we see more partnership announcements in 2026?

What we're learning is the evolution of our partnerships usually start off as an R&D effort, a development effort to see whether our technology can interface with theirs. And so while we did have 2 early partners that were very far along with their technology that just looked for us as a bolt-on to amplify what they've done, one being Mendaera, which is a vascular robotics company, wicked cool, if you see how that works, and the other is Sonic Incytes, that has a fatty liver technology, and they wanted us to help them with one of their use cases. But those were amplifications of their thought, whereas some of these others are earlier where they're really building them into the core architecture.

So Midjourney had been a partner of ours since November of '24, December of '24, but we never talked about them because they purchased our technology, they bought a bunch of chips. They purchased the licenses to our software, which will let them tune the chips, and then they went on to go do their work. And then just about a year later, they made a decision that they are on to something, and they wanted to license it exclusively. And it's that exclusive license that turned into the commercial opportunity. Now we do have other partnerships that are occurring, and we are onboarding more that we are working on. But I don't think, given the early nature of all of this, that we'll be press releasing the individuals unless they decide they want to. But I just think it's important during that development cycle for us to maintain their confidentiality.

Can you share just -- to help understanding just with the pace of these R&D initiatives with partners, where you stand just in terms of potential -- I don't know if you can quantify it or give some qualitative details, but just where you stood in early '25 versus where you stand today?

So I think the cool part about it is we're not making, at this stage, hardware changes. We're not modifying our chip. So the benefit of that is they're taking our existing chip and then they're taking our software, which allows them to modify the settings and program the chip. So they very quickly -- even before we sign a partnership, we may do a lab with them and show them what we can do, that we can meet -- we can be within the boundaries of their use case before they onboard. So a lot of this is software development and software moves quickly.

If we had to then create this whole -- if we had to create a new chip, then that's -- you know how long that would take, and we haven't had to get there yet. But that's the other part of this that's so beneficial to Butterfly is all of this development is making Butterfly better and smarter. And even if we license a certain -- say, we give exclusive license to a certain subsegment of the market, we have access to that technology for all the other segments. So it allows us to be more and more capable. And I think Embedded becomes a significant amplifier of value for the Butterfly shareholder.

Right. And we've gotten a couple of questions just around IP access to patents, the moat that Butterfly has created. My understanding is that there is a moat, and we also have this partnership with the Taiwanese semiconductor manufacturing company that is unique and also create the moat in its own right. But maybe help us think about that competitive moat, both on the manufacturing side and on the IP channel.

You really know until you're tested, and I don't want to tempt anybody to test us. But I mean we do have 600 patents. And there's a lot of know-how in the development of the chip. I don't -- we clearly couldn't have gotten this far without TSMC and they're extraordinary valuable partner. And whenever we move the next chip to get taped out at the fab, they take a whole process of -- and depending upon some of the things we're working on, they may get in with us a little bit earlier, but they don't bring a chip into the fab until they absolutely believe in it and know that it's going to work.

So it cost us $300 million to make our first chip. And then it's been -- we've had the leverage from that. And we just pushed our 4th generation to the fab and all of our engineers are actively working on the 5th generation that has a whole new AI capability to it. So I think that's -- there's a lot of people who will make certain subcomponents. People make CMOS wafers, people make MEMS wafers. But we've been able to marry the -- with a CMOS in a whole tech stack with our software that I'm sure some people can try to replicate, but they have to do it in a different way than we've done it.

We've got a question from the audience, Michael.

There are a handful of other producers of handheld ultrasound. There's 2 names that come to mind EchoNous and a company Exo, I think they call themselves Echo. Two questions, one, have you assessed whether they [ infringe their ] IP; and then two, they didn't enter your mind in marketing and development products?

Sure. So you're talking about both EchoNous and Exo. EchoNous is a company that the founder of SonoSite founded. So it's a very nice technology. It's based upon legacy piezoelectric. And so it's very much a classic legacy ultrasound company. Great company. I think they're probably competing more with the bigs than we are since they're really focused a lot. They have handhelds, but they're focusing on carts right now a lot. So that's not -- we don't really encounter them on a day-to-day basis. Exo makes a piezo-based chip. And so they're the only other company that's -- at least they've tried to commercialize a chip. They're on their 1st generation for everything that we know. They commercialized a product called Iris, unfortunately, for them, right at the time we launched iQ3. And our iQ3 was just next level to where that performed. And it hasn't had the type of traction in the marketplace.

But we think about them a lot because we are a CMOT-based chip and they're a PMOT-based chip. And there are different physics between the different MEMS of those chips. And so we have to be very respectful of what they could potentially create in the future. So yes, we do think about them. We don't see them in the marketplace, but we know they're there. They just received $100 million cash infusion from Samsung almost a year now, maybe less than 9 months ago. So you have to give tremendous respect to Samsung. You have to give tremendous respect to them. So I'm sure they'll do something at some point, and we do think about them.

We did spend a big clip of this discussion on Butterfly Embedded, but it's appropriate considering everything that's going on in the last couple of months. But thanks to Mike, we'll pivot over to point-of-care ultrasound segment in iQ3, Butterfly iQ3 and the pipeline. But maybe just to talk about the market and your handheld focused business. It sounds like that there's sustainable double-digit growth in play currently, but there are some building blocks that are creating a foundation for potentially an inflection in the market. So maybe talk about kind of market growth assumption rates, how Butterfly is positioned there, which is very well positioned is our view? And then are there -- can this market inflect in -- over the next 3 to 5 years?

Well, I've joked many times like Butterfly kind of had a fast start and then had to go through some kind of resetting of expectations. But I've always felt our founder and I felt the prior CEOs before me have always had the right strategy and always had the right vision that empowering every doctor, every nurse in the world with a device is an absolute possibility. But I think that probably the biggest difference I've had is simply in the thought that it takes 10 years to become an overnight success. And in health care, you just have to be mindful of changes in practice and workflow and having to get into how you change everyday clinical practice of medicine.

Well, that 10 years is coming up. And I think it's happening, to be quite honest with you. Right now, about 80% of medical students to -- one way or another are getting trained on Butterfly. So we've already won the game. The die is cast. It's a fait accompli. Ultrasound will be a part of clinical practice. I just don't want to wait 20 years for it to all of a sudden fully inflect. So now all of our marketing practices are on how and what do we do to compress that back. So we're not just waiting for the next generation, but we're helping this current generation evolve. So we've done everything from education tools. We have our own University of Phoenix type training with Butterfly University. And then we have our AI tools with ScanLab.

So we've done everything. Now we've launched Garden that we're going to have 20, 30, 40, 50 new apps over the next 5 years that come out that make it easier to do ultrasound. And we're starting -- and we've now invested in our next generation of our Compass, which is our enterprise SaaS-based software that allows us to help manage our data and manage quality and get patient records, images and reimbursement files into their system. I'm feeling that hospitals are starting to wake up. I remember a year ago, we went to a conference and sat with 14 CMOs and asked them about their point-of-care ultrasound program or their POCUS program. And they said, what's POCUS? Chief Medical Officers. What's POCUS? But we're actually starting to see hospitals starting to coalesce around creating infrastructure, creating guidance, creating standards because about 70% of the scans that are being done are uncaptured, are not reimbursed, are not compliant because it's not completing the patient record.

So this is happening. People are waking up. And I think the business is probably going to be pretty lumpy because we're going to see a health system all of a sudden go and it will be a big order. And will that happen next? I don't know. Let's see. But I'm starting to feel that point-of-care ultrasound is going from being nascent to now starting to get into the 4 walls of the hospital. They're starting to think about it more. And all the big companies want to portray point-of care ultrasound as a cart. But over the next 18 months, we're going to come out with a new technology that we've talked about that's going to change the concept of every doctor and every nurse having their own device and change the concept of putting one-to-one in a hospital.

So I'm really bullish. I think our core business is going to -- we will see basically a double-digit growth, I think we believe. And we think that the investments -- we just had our POCUS Innovators Forum, which is 60 of the top POCUS leaders that came in, and we had a conference. And literally, what they're talking about that they need for the future is all of our road map that we are working on. So the convergence of the thought and where we are for the future, we think, is right online.

Great. And can you just build out on that product you're referencing that within the next 18 months that you're going to introduce that's going to potentially be transformational in terms of...

The product is called iQ Station. And basically, what we see in the hospital are carts and there's carts everywhere, and there's carts for this use case, that use case. You show it to the cart and you kind of have to deal with the experience that, that cart gives you. But everyone today has their own phone. And -- but let me just ask, if you were using your phone, would you do -- would you create a PowerPoint presentation on your phone? Would you create an Excel on your phone? No, you don't do that. You go back home, you sit at your docking station, you got your 2 screens and your laptop and then you have this environment where you're going to sit there for a long time and do that kind of detailed work.

Well, on the one-to-one side, everyone is going to have their own device, but it's going to be amazingly powerful, and it's going to have the image quality that you need to be able to do the types of scans that you need. But when you go inpatient to the hospital, you can actually dock the device. And then your whole experience, all your presets, all your patients, all your workflow now gets expressed on this device. And the device may even know who you are before you walk in the room and when you walk in the room and ask you, do you want to pair. And so it takes your Butterfly experience, but gives you -- takes your Butterfly data and know-how and gives you an inpatient experience. And so this will be less expensive than the carts out there, and it will be -- it will allow us to tap into the capital equipment budgets of all the hospitals to completely, summarily take out all the low-end carts and turn them into one-on-one model.

Appreciate that. That deepens our understanding of that initiative. And John, sorry, I've been leaving you on the edge there, but I'd love to just touch on 2026 guidance, and just some of the different drivers that get you to the revenue growth and anything you want to share just on the gross margin levels expected in 2026?

Yes, sure. So the guidance we put out, just in case folks aren't paying attention, for the full year is $117 million to $121 million. That's up 20% to 24% year-over-year. We do believe the core business should grow at high single digit to low double digits, and the rest of that will come through embedded. We got gross margin to north of 67%. We're comfortable there. The embedded business is throwing off a really solid margin level overall.

That said, we also gave Q1 guidance. Q1, we do tend to trade down a little bit. Certainly, we'll be up year-over-year, but from a sequential perspective, we expect to be down. It's typically a weaker quarter. Also, we have 401(k) resets. We have other tax -- income tax based things that reset as well. So from that perspective, we guided to a negative 8% to 10% -- negative 8% to negative 10% net loss and will likely be the worst quarter that we have from that perspective. But what we're -- just more generally, what we're trying to do within the business is really build operating leverage. We're in a really strong cash position. We have over $150 million in cash. We can fund the business going forward.

In addition, we're working ourselves to be a cash flow positive company over the period of time, we've been out there talking about by the end of 2027. I think we're certainly on track for that. If you look at the guidance overall, just to fine-tune it a bit, give or take, revenue up $20 million in absolute dollars. Our net loss improving by about $5 million. So we are continuing to also invest in the business. And with our cash position, if we see opportunities to accelerate the top line growth, we're going to be in a position to do that.

And just thinking about the core handheld business and that enterprise channel that you guys have been attacking, pipeline seems to be flourishing. There's been some puts and takes on the capital equipment spending environment in 2025. It's hard to forecast how that's all going to pan out in 2026. But is that a fair assumption that just the enterprise pipeline in terms of potential customers that are moving forward potentially with contracts and either on the Compass software side or the handheld hardware side or a combination of both is in a better place than it was at the beginning of 2025?

Yes, it is. Yes. We feel -- let's just say it's back to normal plus another year of market awareness and involving. So we're pretty excited about where our core business is, and we're pretty pumped about '26.

Excellent. Well, I think it's a good place to stop. There's a lot more to address, and we're going to do that down the line as looking forward to tracking continued progress throughout this year, and maybe we'll hear from Midjourney and get some more learnings on this Butterfly Embedded partnership...

You we'll be the first phone call. I'll call you once it's up [indiscernible]. You deserve that.

Thank you, guys, for participating again this year. Have a great rest of the day. Good luck with your meetings this afternoon.

Appreciate you, my friend. Thank you for having us.

Thanks, Josh.

Hello, everyone, and thank you for joining the Butterfly Network, Inc. Q4 and FY 2025 Earnings Call. My name is Gabrielle, and I will be coordinating your call today. [Operator Instructions]

I will now hand over to your host, John Doherty. Please go ahead.

Good morning, and thanks to all of you for joining our call today. Earlier, Butterfly released financial results for the fourth quarter and full year ended December 31, 2025. We also provided a business update. The release, which includes a reconciliation of management's use of non-GAAP financial measures compared to the most applicable GAAP measures are currently available on the Investors section of the company's website at ir.butterflynetwork.com.

I, John Doherty, Chief Financial Officer of Butterfly, along with Joseph DeVivo, Butterfly's Chairman and Chief Executive Officer, will host the call this morning.

During the call, we will be making certain forward-looking statements. These statements may include, among other things, expectations with respect to financial results, future performance, development and commercialization of products and services, potential regulatory approvals, revenue attributable to embedded collaborations through revenue share, chip purchases or otherwise and the size and potential growth of current or future markets for our products and services.

These forward-looking statements are based on current information, assumptions and expectations that are subject to change and involve a number of known and unknown risks, uncertainties and other factors that may cause actual results to differ materially from those contained in the forward-looking statements. These and other risks are described in our filings made with the Securities and Exchange Commission. You are cautioned not to place undue reliance on these forward-looking statements, and the company disclaims any obligation to update such statements.

As a reminder, this call is being webcast live and recorded. To access the webcast, please visit the Events section of our investor website. A replay of the event will also be available on this page following the call.

I would now like to turn the call over to Joe.

Thanks, John. Good morning, everyone, and thank you for joining our fourth quarter and full year 2025 conference call. I'm pleased to announce that our fourth quarter revenues came in at $31.5 million, growing 41% year-over-year. It's also the first quarter in our history where we realized positive operating cash flow, driven by upfront payments from our Midjourney deal.

We are now nearly 2 years into our strategic growth plan that we introduced in early 2024. The message then was clear: execute with focus and financial discipline, strengthen our core franchise and unlock the full value of our semiconductor ultrasound platform beyond handheld devices. We're doing just that, and we now have material contribution outside of our core POCUS business.

So earlier this month, Jim Cramer was asked a question about Butterfly Network on Mad Money. And he said, I quote, "It's a very competitive market, and I don't want to be in that part of the medical device group. Too hard for this guy, just too hard." Well, I'm a big fan of Jim Cramer and, like you, watch CNBC every day, and he's right about one thing: traditional medical devices are a tough sector.

But Butterfly is not a traditional medical device company. We were never just a handheld company shaking at the knees of big ultrasound. We are David, faithful and committed to our cause, not backing down, regardless of how large our opponent is. We are the disruptor. We are playing our own game with our own rules and our own playbook. We are digital, eating analog's lunch and plan to revolutionize imaging with digital ultrasound everywhere and images taken. Goliath just doesn't know it yet.

So we are changing ultrasound imaging with our differentiated chip platform while deploying secure cloud software, AI tools and mobile applications and are the only health care company in the world over the last 7 years to receive an Apple design award. We are not a medical device company alone. We are a transformative semiconductor-based ultrasound company.

As you're all aware, the strategy we introduced has 3 core tenets: accelerate our core focus business with an enterprise-ready cloud-connected imaging solution; execute strategic initiatives to reach new care settings and enable entirely new applications, namely HomeCare and Ultrasound-on-Chip co-development; and then lastly, deliver an R&D road map that sharpens our technology edge with next-generation chips and new form factors getting more and more powerful and capable.

So let me walk you through the progress in each of these. In core POCUS, revenue grew 15% year-over-year on top of a 35% growth in the fourth quarter of last year, which, as you recall, was a record product launch year. We closed a second large system-wide enterprise deal in the fourth quarter of '25, and continue to deepen medical school and enterprise relationships.

I'll let John cover the financials in more detail, but we feel much better about macro trends and the deal cycle than earlier in the year and expect to carry that momentum into 2026.

The fourth quarter Compass AI launch is a big tailwind for our enterprise strategy. Enhancing our enterprise solution helped refresh existing accounts, continue attracting new customers and drove more than 50% growth in our enterprise pipeline since launch. At the same time, we've been building what we believe is one of the most secure cloud infrastructures in health care.

In the U.S., we most recently achieved GovRAMP and TX-RAMP, and FedRAMP is anticipated in the coming months, meaning we'll soon be even more qualified for cloud deployments in now the world's largest government market.

Internationally, we see meaningful opportunities in 2026 as we open markets in South America and continue expanding across Middle East and Asia, alongside a renewed momentum in global health. We've surpassed 1,000 NGO partners worldwide and many expect to increase activity in 2026.

We also continue to make progress in our Butterfly Garden partners. And as you know, HeartFocus became the first FDA-cleared app in 2025, and we expect additional partners to reach that clinical milestone in 2026.

Importantly, we announced plans to release our proprietary beam steering API in the first half of this year. The API will open 3D imaging capabilities that have been reserved for Butterfly products only. Because our beam steering is fully digital on a semiconductor chip and not mechanically driven like other legacy systems, we can uniquely extend those capabilities to partners. The goal is to advance what they can build, expand the AI ecosystem and help accelerate ease of use. We're excited about the new opportunities this welcomes.

So moving to the strategic growth engines. We've built a home business determined to accelerate the use of POCUS devices where the patient is, expanding the reach and empowering nurses and other clinicians to perform scans using AI tools. I believe we will reach a commercial agreement in '26, allowing HomeCare to enter its commercial phase. This is a powerful new channel that reduces hospital readmissions, lowers costs and expands Butterfly's reach beyond the hospital. HomeCare is a reality, and we think it can be a real growth driver starting to add revenue in late '26 and into '27.

We also ventured out to develop a new business to make our core Ultrasound-on-Chip available to companies with large new market opportunities that are not competitive with Butterfly's POCUS business. That was previously known as Octave. Well, to start the new year, we've officially sunset the Octave name and have centralized our semiconductor platform strategy under Butterfly Embedded, reminiscent of the Intel Inside model.

What began as an adjacent effort is now emerging as a foundational business, enabling other category-defining innovators to build entirely new application on Butterfly's Ultrasound-on-Chip technology.

The big news for the fourth quarter, though, was signing Midjourney deal in November. And as you can see today, it contributed $6.8 million of revenue in the quarter and was a key driver of our 41% year-over-year growth.

Midjourney is an independent research lab, pushing the boundaries of generative AI and human imagination. Their scientists have envisioned a breakthrough new application for ultrasound that we believe the world will be learning about very soon. This partnership combines Butterfly's core Ultrasound-on-Chip technology and imaging software with Midjourney generative AI compute power to do something very special with ultrasound.

So this deal is more than revenue. It is validation. It's a foundational step towards Butterfly's reinvention into a platform company. The initial partnership and revenue contribution is pre-commercial. So based on the plan that Midjourney has shared with us, once commercial, we believe there is meaningful additional revenue opportunity for Butterfly in the form of chip sales and revenue share on top of the licensing fees that will continue through the deal. We expect this to happen in the outer years of our 5-year plan. And if it occurs, it could represent meaningful progress towards our goal of reaching $500 million in annual revenue by 2030.

Beyond Midjourney, we signed an additional Butterfly Embedded research share partner this quarter and expect another one shortly while managing an active pipeline of some of the largest technology and health care companies in the world.

So before I look ahead to our R&D road map initiatives, I'll turn it over to John to discuss the financials. John?

Thanks, Joe. I'm very excited to have joined Joe and the talented team here at Butterfly, which includes Megan Carlson. Megan, who you all know, did a great job in the interim CFO role and has been an awesome partner to work with along with the rest of the leadership team. With about 3 months behind me, I want to open with a few comments and my thoughts on Butterfly overall.

I joined Butterfly because I was excited by the incredible potential this company has going forward to provide a valuable and meaningful experience to our customers as well as to create value for our shareholders. I was also impressed by the strength of the leadership team and the passion and commitment demonstrated by employees across the company. This is a company I am proud to be a part of.

Butterfly certainly navigated a few challenges in its early years as a public company. However, the company has made the necessary and difficult adjustments, including improving its operating efficiency, choosing the markets it can win in, selecting the best ways to leverage its IP and technology advantage and allocating its resources towards higher ROI opportunities and markets.

No doubt, based on these actions, Butterfly is a stronger, more agile company today. With continued steady execution, the company has the potential for significant core growth in an incredibly meaningful domain through its POCUS and related Compass AI software and Butterfly Garden business, and we are well positioned to gain outsized share in the markets that we compete in.

We also have a significant opportunity to leverage our core platform of Ultrasound-on-Chip into other nascent and disruptive markets, consistent with our strategy, through Butterfly Embedded as the Midjourney partnership demonstrates.

With that, let's move on to a few highlights for the fourth quarter, including a record level of revenue for the quarter and exceeding the high end of our total revenue and adjusted EBITDA guidance ranges, a record level of quarterly probe sales, significant improvement in adjusted EBITDA margin, driven by our revenue performance and continued financial discipline, the lowest annual cash used in the company's history, and we generated positive free cash flow in the quarter and the execution of the Midjourney contract, which helped to solidify and amplify all of our efforts to drive a new wave of ultrasound-enabled platforms and use cases through Butterfly Embedded.

Now let me move on to our results. We had a strong fourth quarter 2025 with revenue of $31.5 million, the highest quarterly result in the company's history. This represents a 41% increase year-over-year for the quarter. The increase is driven by increased revenue from Butterfly Embedded as well as a 27% increase in year-over-year volume in our core POCUS business with continued penetration of the iQ3 in all markets.

The 41% year-over-year growth is significantly higher than the at least 17% year-over-year growth that we put out in January. We did better in the core business and realized more revenue from Butterfly Embedded, given the work we have already performed relative to the Midjourney contract. This was the primary reason we use the at least language at the time as we were still finalizing the accounting treatment. Our results and the guidance I'll provide later in the call for first quarter and full year 2026 all include updated financial expectations from this contract.

Breaking things down between the U.S. and international channels, during the quarter, U.S. revenue was $26.8 million, which was 55% higher year-over-year, driven by revenue from Embedded as well as strong demand in the core business with unit sales up 44%. Total international revenue decreased by 6% year-over-year to $4.7 million in the fourth quarter. While sales of the iQ3 in the quarter were up 42% year-over-year, sales of the iQ+ were down 79%.

Breaking our revenue down between product and software and other services, product revenue was $18.1 million, an increase of 23% versus the fourth quarter of 2024. This increase was driven primarily through growth in volume across all channels, with U.S. health systems, e-comm and vet leading the way as well as higher average selling prices in international markets with the higher mix of iQ3 sales.

Software and other services revenue was $13.4 million in the fourth quarter, up 76% year-over-year. Software and other services mix was 43% of revenue, increasing from 34% in Q4 2024. This increase can be attributed to the significant step-up in revenue contribution from Butterfly Embedded in the quarter related primarily to the Midjourney partnership. I'll talk more about this as well when I come to guidance for the first quarter and full year 2026.

When looking at the full year 2025 versus 2024, total revenue increased 19% to $97.6 million. This was driven by 2 areas. First, growth in our core POCUS business with both increased volume and a higher average selling price, driven by an increased mix in the sales of the iQ3 and strong performance by U.S. sales, our vet channel and international distributor channels; and second, growth in our emerging Butterfly Embedded business, primarily from the execution of the Midjourney contract in November of last year.

Moving on to gross profit. Gross profit was $21.2 million in Q4 2025, a 55% increase as compared to the prior year adjusted gross profit of $13.7 million. Gross profit margin percentage increased to 67% from 61% in the prior year period. Gross margin percentage was positively impacted by the higher-margin Butterfly Embedded revenue and lower software amortization.

Moving to EBITDA and cash. For the fourth quarter of 2025, adjusted EBITDA loss was $3.2 million compared with a loss of $9.1 million for the same period in 2024, an improvement of 65%. For the full year 2025, adjusted EBITDA loss was $26.5 million compared to $38.9 million for 2024, an improvement of 32%. The improvement in adjusted EBITDA loss for both the fourth quarter and full year was driven by contribution from higher-margin revenue and continued financial discipline reflected in our lower year-over-year payroll costs for the quarter and full year.

This improvement in adjusted EBITDA and continued financial discipline has led to a cash and cash equivalent balance, including restricted cash at year-end of $154.5 million and the use of cash in 2025 of $19.4 million, excluding the funds from our offering last year. This compares to a use of cash of $45.9 million in 2024, an improvement of $26.4 million. We also had positive cash flow of $6.3 million in the fourth quarter.

These results demonstrate that we are very well positioned as we move forward to continue to invest in the business areas where we see significant opportunities for additional growth and disruption, including: expanding our core POCUS business and penetration of Compass AI as a core operating system for health systems; continuing to enable third parties to build tailored ultrasound AI solutions in Butterfly Garden to provide deeper and expanded access to our platform; expanding our HomeCare business following the execution of our anticipated first commercial agreement, which we believe could be finalized in the first half of 2026, as Joe mentioned earlier; enabling a new wave of Ultrasound-on-Chip-enabled technologies through Butterfly Embedded; and continued AI and semiconductor innovation with the development of our fourth-generation chip.

Before turning to guidance, I want to update you on the general macroeconomic environment relative to Butterfly. While there were some concerns in 2025 relative to the government shutdown and impact on the FDA's processing of fee-based submissions, regulatory processing delays and delays in customer purchasing decisions, we have managed through this, and it is very much behind us. Our fourth quarter results are indicative of that, and we were able to close some of the larger deals in the pipeline and still have a number that are active.

I would now like to turn to our outlook for the first quarter of 2026 and for the calendar year ending December 31, 2026. In the first quarter, we expect revenue in the range of $24 million to $28 million as we ended the fourth quarter somewhat higher than expected and the first quarter is typically a slower quarter for the company due to seasonality.

As is typical for Q1, adjusted EBITDA is impacted from expenses related to payroll tax and 401(k) reset as well as our national sales meeting in POCUS Forum, which took place in January. As a result, we expect slightly higher expenses and higher adjusted EBITDA loss in the first quarter relative to the remainder of the year in the range of $8 million to $10 million.

For the full year 2026, we expect revenue to be between $117 million and $121 million, an increase of approximately 20% to 24%. We expect our adjusted EBITDA loss to be between $21 million and $25 million. Our guidance for adjusted EBITDA includes investment in key areas to support continued innovation as well as our emerging embedded business as well as the impact from tariffs initiated in 2025.

In summary, we had a great quarter, a record quarter. We closed the year out strong. We exceeded expectations for revenue and adjusted EBITDA, and we are very well positioned going forward. As our 2026 full year guidance indicates, we look forward to continued growth this year and beyond, and our overall outlook on the business is brighter with this past quarter reinforcing our view.

In 2025, we believe the company enhanced its position in the core POCUS markets and can continue to gain share in 2026 through deeper penetration of existing customers, new customers and applications. We also amplified our Ultrasound-on-Chip platform, and the potential of Butterfly Embedded with the licensing partnership with Midjourney. We did all of this while continuing our intense focus on driving operating efficiency across the business and return on investment.

As I said upfront, I'm excited to have joined Butterfly late last year, and I'm excited about what is ahead for the company in 2026 and beyond.

Now let me hand it back to Joe for some closing comments.

Thanks, John. It's an exciting time for Butterfly. We're in a strong position and are delivering on all fronts. The future is even brighter.

So before closing, I'll touch briefly on R&D. We recently spent 2 days with 60 leading POCUS thought leaders at our annual POCUS innovation forum and the alignment was unmistakable. The goals, needs and future state vision shared was directly in line with where we are taking our portfolio. That level of synergy only strengthened our conviction that we have a path to enable every doctor and nurse with powerful, affordable compact imaging and can meaningfully unlock enterprise adoption.

As I mentioned on the last call, our fifth-generation P5.1 chip was moved to production by year-end. We're excited in achieving harmonics and delivering a new level of imaging for handheld ultrasound. This will open additional subspecialties and support our enterprise selling efforts in 2027 and beyond.

And then next up on our chip road map is Apollo. This new chip architecture is now receiving the full attention of our engineers and will deliver 20x the current data rate and compute performance, ushering in a new era of digital imaging and AI. We are already working with several Butterfly Embedded partners based on the capabilities of this platform.

Now I'd like to close by sharing exactly why I'm so excited about Butterfly's current chapter. We are pioneering semiconductor-based digital ultrasound everywhere it's needed. In the traditional ultrasound market, that means continuing to lead in point-of-care solutions that truly meet the demands in and out of the hospital, empowering doctors, nurses and caregivers, expanding new subspecialties, strengthening our enterprise road map, moving imaging to the bedside and beyond it.

But it's also bigger than that. Ultrasound is being researched and deployed in brain therapy, continuous monitoring, robotics, organ preservation and the list goes on. If you do a simple search of ultrasound in the magazine Nature, just this month alone shows applications in neuromodulation, ablation, antiviral therapy and soft robotics. The real inflection point comes when ultrasound in these use cases move to silicon, when it becomes programmable, power efficient, AI native, scalable. That's the common thread.

Whether it's a Butterfly handheld in a clinician's pocket, a new form factor on our road map driving enterprise or home adoption or a partner building something novel with our technology that we can never do ourselves, it's the same core technology, the same architectural advantage.

We are not just building devices. We are opening markets. We are enabling a new age of digital ultrasound imaging and sensing, all in one strategy. This is Butterfly. We're not just a medical device company. We're a true disruptor building the ultrasonic backbone of the AI era and one of the most exciting technology growth stories in health care and technology today.

So with that, operator, please open it up for questions.

[Operator Instructions] Our first question is from Josh Jennings from TD Cowen.

Impressive to see all the progress in so many different channels, gives us a lot to ask about. It's challenging to choose one of those lanes, but I wanted to start with Butterfly Embedded. Just wanted to better understand, and you may not be able to share, but any details around potential timing of the technology offering that Midjourney is going to put forward, implementing Butterfly's semiconductor and Ultrasound-on-Chip technology? And any help just thinking about the cadence of revenue contributions from this partnership over the course of 2026 and what's embedded in guidance?

Well, I'll let John answer the second part of that in a moment. Let me just get to the first. So I think it's -- we're probably going to see something in the relative near term for Midjourney. We -- one of the things about a Butterfly Embedded program is we are enablers for our partners. This is not our business. This is their business, and we do everything we can to amplify their business. And so I don't want to get out ahead of them. And all I want to do is do everything that we possibly can as a partner to support them. That said, I think they probably want to get this out sooner rather than later. And the moment they do, we'll do the best to show you how it translates into Butterfly.

John, do you want to take the second part of that?

Yes, sure. Thanks, Joe. Appreciate it. So when you look at embedded, obviously, and as Joe touched on during the opening remarks, Midjourney kind of lit it on fire, if you will, with the contract we signed at the back end of 2025. And we're very excited about that contract. But there's also a number of other partners that we're working with in embedded.

So while right now, embedded in our revenue is -- there's a big contribution from Midjourney, we do expect other contributions from other partners that we have in there. And we're really excited about what that part of our business can be as we discussed at the back end of 2025.

Relative to Midjourney, it's a $74 million contract, as we mentioned. But there's different components of it. There's the upfront payment, there's annual license fees and then there's milestone work that we do that ultimately allows us to do the -- take the revenue ultimately quarter-by-quarter, if you will, year-by-year. We expect to get a good amount of contribution throughout 2026 into 2027 from the contract.

And then in addition, when they commercialize, ultimately, as Joe mentioned upfront as well, there's an opportunity for chip sales as well as for a revenue share as part of their business. So ultimately, really excited about the overall contribution from Midjourney, but embedded is not just Midjourney.

I appreciate that answer. And it kind of leads into my next follow-up is -- I mean our understanding is that the team was focused on the Midjourney partnership and development work and then locking that in, in 2025, but there were some other potential partnerships on the periphery that sounds like you guys are engaged and moving forward. But any help thinking about the Butterfly Embedded pipeline that you just referenced?

Yes. I mean -- so we're talking to a lot of people, as I mentioned, some very large organizations and also a lot of start-up organizations. And kind of the way it works is we saw a contract in the 8-K with Midjourney in November of '25. But they had been an embedded partner for over a year prior.

And so the way it kind of works is people will buy a license to our software and then they'll start buying chips and they'll do some research. They'll do research as far as how does -- does the chip meet their performance that they need, how does it integrate with their systems and then -- and they do a bunch of work.

And whenever they're done and ready, if everything worked out well, then it will turn into a commercial agreement. They'll come back and say, okay, well, we want to now do X, Y and Z, and that's kind of how the journey happen. So there's really no formula. It's just -- we have, I think, now 8 or 9 embedded partners today, and the pipeline is pretty large.

But it's also -- it's -- we're not selling off-the-shelf product. We have to take our product and we have to do labs and show them how it can fit with their tech and whatnot. So it is a bit of a sales process. But do we have another Midjourney in there? I hope. There are some very big opportunities, people are going after massive opportunities actually and we're going to continue to add people into the research phase of our relationships and do everything possible to meet their needs and get them committed to a commercial phase like Midjourney did.

Excellent. And maybe one question on the HomeCare franchise. And it sounds like you've made progress in this first commercial partnership. I mean what steps are left before that commercial effort kicks in? And maybe just remind us of the home POCUS market opportunity or the revenue opportunity for Butterfly in this channel.

Thank you, Josh. Appreciate them. So yes, I think we're pretty confident that this is going to get to a commercial conclusion. As John said in his notes, we think it will probably happen before middle of the year, which will give us the second half of the year to start getting it ramped up. So it'd probably be nominal revenue in '26, but start contributing in '27.

And so once we get it started and we have a line of sight and we're able to give a better shape of what the impact is, we'll tell you as far as incorporated into our guidance. But in aggregate, I think I've mentioned in the past, just this one use case we're working on, if it became a national use case, it could be a $40 million, $50 million revenue opportunity if everything went the way that we saw it.

So -- and that's -- we -- when we think about the amount of chronic care patients in nursing homes and also the amount of patients that are in the home who are being cared for by nurses, but then, as they get sicker, they need to be moved -- they need to have an image. And so they have to go to the hospital, whether it's by ambulance or by some other transport, be brought up into radiology and then have, for example, a cardiac echo or something else.

Empowering caregivers to be able to take images where they are, is efficient, is low cost and has high impact on being able to quickly give a diagnosis for a patient so their meds can be modified to make them healthier and keep them into the home or keep them wherever they are, and that is literally -- every facet of health care is about being where the patients are, being where people are and helping them get healthier, faster in the lowest cost environments.

And so our HomeCare business is basically an extension of our POCUS business. It's not a third business. It's an amplifier of POCUS because ultimately, everyone will figure it out, and they'll use our stuff to do all of this using the AI apps, et cetera. But home is kind of a catalyst for us to take a lot of the workflow, the logistics and the ramp-up and the learning curve out.

So we think -- I think I've mentioned before that we think the home business can be bigger than POCUS and the opportunities are certainly there. But we have some -- we have to prove it, and we have to get some more -- our pilot was excellent. We need to get the first few states up. We need to prove what we're doing there, and then we need to prove we can go to a larger geography.

And so it's a process, but it's a big opportunity, and it will teach core health care providers and payers that point-of-care ultrasound with Butterfly is a significant way to reduce costs at the exact same time of improving quality and care for patients.

Our next question is from Chase Knickerbocker from Craig-Hallum.

This is Jake on for Chase. I was wondering if you could -- maybe, John, if you could give us any more color on the macro environment, if you guys are seeing any of the same trends that affected 2025 and how they're abating now in 2026.

Yes, sure. And I appreciate the question. So as I touched on with the opening comments, I'm not going to say everything is behind us, but certainly, we feel a lot more confident going forward. Irrespective of the Supreme Court's action relative to tariffs, it's still a very, say mildly, an unpredictable area, given what happened post that. So we do still expect to have a bit of some downward pressure, not a lot, but some downward pressure from tariffs.

We did sign a number of contracts late in the year. So we see that things have been opening up a bit there. We still have a number of good-sized deals in the pipeline. I'd say timing did move out kind of the back half of last year as the company had touched on when they did third quarter results, but we are seeing some improvement there. We're starting to see some things move. And ultimately, as I mentioned, the company, we're managing through it, and we don't expect it to have -- except for the tariff component, we don't expect it to have a lot of impact on the company.

Great. I appreciate that color. And then apologies if I missed this, but is there any more color on the opportunity with medical schools going into the year as well for the one-to-one partnerships?

Yes. So medical schools, I think, should be a pretty strong contributor to us this year. The second quarter of the year is the big medical school quarter when we put them all in. We have a lot of conversations for one-to-one buys. We did get a bit of a pause last year when the decision to cap student loan debt occurred, and so it caused people to pause.

But what's so exciting about what we're doing is there is such a strong commitment for new medical students to learn ultrasound. And they're actually now starting to choose where they go and the types of programs they get into, based upon the commitment of building that competency.

Kids that are graduating today and are going into their residency completely distinguish themselves when they have this capability. Even their attendings who are working with them are kind of surprised when they see what can be done by a resident with ultrasound skills.

So there is not a lot of daylight out there. This is not a new thing. Students want Butterfly. They're trained on Butterfly. They want that experience. And then when we build the brand equity with these students, they go into their residency and then into practice with that loyalty, with that understanding, and we're looking to grow with them.

Especially with our whole chip platform improving and getting more and more powerful and then more apps coming in over the years, we're going to grow with them through their career. So we start off with medical schools. Medical schools will be a very positive contributor in 2026.

Our next question is from Andrew Brackmann from William Blair.

I'll echo Josh's comments. Certainly, a lot of progress here and a lot to dig into. So Joe, maybe with that in mind, I'd like to start a little bit higher level. I mean it's sort of hard to miss the underlying transformation going on here towards Butterfly Embedded. Can you maybe just sort of talk to us about why now for this transition from med devices towards embedded and semiconductors type of business? And then just as you sort of think about organizational focus, perhaps are there any sort of changes in terms of your view on POCUS versus the opportunity with what's embedded?

That's a great question, Andrew. And I actually -- I probably should have e-mailed that to you and planted that in because it's -- that's a very appropriate question. We're not trying to pivot away from POCUS. We want to have every doctor and every nurse in the world have a device that allows them to help diagnose their patients sooner, easier, low cost, where they're at. We will never move away from that vision.

But in accomplishing that vision, Andrew, we solved the problem. We solved the major problem in technology where we can digitize. We can digitally control the ultrasound image. And we can control it in a way that is much more than what medical ultrasound uses with multiple planes and depths, because we control 9,000 sensors in a rectangle, which allow us to send out information, send out sound a certain way and listen to sound a certain way.

And so what we solve for ourselves for point-of-care ultrasound is a major solution for many other people out in the marketplace. There is this whole concept that AI and the human body are going to merge together. You can google it. There's a whole sci-fi fantasy out there. And there -- if you even think of brain computer interface, Neuralink today is putting wires into the brain. And there's this thing in DCI called the butcher factor, which is how many neurons do I have to kill before I get to the neuron I want to talk to.

Well, one of the beautiful things about ultrasound is -- and about Butterfly is we don't have a stagnant beam that sits in a fixed position. Normally, ultrasound is like a flashlight. You have to move the beam in order to capture the image. We can actually, in a fixed location, move our beam and scan. So if you implant something in the brain, you're going to be able to communicate and see and listen to 25% of the entire brain or the entire lobe.

So it's being seen today that ultrasound is potentially a portal between AI and the human body. And just read Nature. And you'll see all these articles about brain-computer interfaces and you'll hear about ultrasound, you'll hear about neuromodulation, you'll hear about all different types of -- even if it's therapeutic, people are going to be looking at the brain signals through vasculature and how vasculature flows through the brain. And ultimately, you can potentially heat it and put energy.

So this is kind of happening. It's not like we just did a prospective randomized study, and we're now pushing the market. The market is calling us. They see our 600 patents. They see the excellent technology that we have. And quite frankly, we have been approached, prior to me getting here, by companies asking to partner with them, but they were turned away because the company was trying to focus on point-of-care ultrasound.

So what we've said to ourselves is, no, we've solved the foundational problem. And I'd hate to compare ourselves to NVIDIA because there's no way there's a scale or anywhere to draw a line. But from '93 to 2005, they were a video game company. And they focused on their core markets of building technologies to be better gamers that had to be complex. And then with CUDA in 2006, it opened up their platform for developers who could use these incredible chips for other applications.

And -- or you look at Amazon, they didn't set out to become the leading cloud compute company in the world. They realized that they needed cloud compute. There wasn't a large enough vendor to take their volume. They built it themselves. And when they were done, they had excess capacity. And they said, let's go sell that capacity. And now it's the highest profit-generating part of their business.

Sometimes in business, you solve problems for your customers that benefit others. And we're just simply now allowing ourselves to work with those partners. And the beautiful part about it is that the road map that we set ourselves, so our Poseidon platform, our P5.1 platform, our Apollo platform, we are now not going off and developing new things for other people and losing our focus on point-of-care ultrasound. What we're actually doing is amplifying our current road map, and we're exposing our road map to those customers and those -- to allow them to meet their needs.

So it actually is going to make our ability to serve point-of-care ultrasound better. And the road map that we have for point-of-care ultrasound and what's going to happen with our next launch of our platform with harmonics and P5.1 and what we're going to do with some new things that we'll unveil in 2007, there's so much synergy here, Andrew. It all ties together.

So we're not putting something on the shelf and then focusing someplace else. We're pulling our core tech. We're partnering with people who have real interest of building businesses and have technical challenges that we can solve. Instead of having an analog ultrasound device that -- yes, you have digital behind the lens, but then once you push the signal through the lens, that lens is cut to do one thing.

Our lens can be anything. And developers love the fact that they can see the signal and then change the software, see the signal, change the software. And that kind of infinite iteration cycle is how we will fit our ultrasound into many different types of devices throughout the world.

Okay. That's terrific color. Maybe if I could, just sort of switching lanes here. You referenced the Compass AI and sort of the launch there. Can you maybe just expand on sort of what that does in practical terms for your moat within that POCUS business? What does it give you that competitors don't have today? And then, as you sort of think about defending against some of these potential AI -- other AI disruptors that are out there, how does this sort of expand the moat there?

Thank you so much, Andrew. So we were -- the prior team before I got here, Todd Fruchterman and the team were very visionary on making sure that if point-of-care ultrasound is going to be adopted enterprise-wide, that we need to manage the data. And the systems, the EMRs, the DICOMs, all the current hospital systems were not geared and built to operationalize point-of-care ultrasound.

And so Butterfly became the first ultrasound company that invested in an enterprise SaaS-based software that not only allowed it to collect its own data, push it into the record, push the image into DICOM and set that record up, ready to be submitted for reimbursement with the revenue cycle management software, but we've built the most powerful SaaS-based system out there in order to do this.

And now hospitals, when they put our software in place, they now recognize scans that weren't getting reimbursed, and they increased their revenue and profitability. And so it's just good, good and good.

Now on the other side, doctors hate new things. They don't like to have to go through new steps. They just want to do everything in their EMR. They don't want to have one software for this, one software for that. So getting to an ease-of-use paradigm is essential because the doctors are already asked, and nurses, to do so much in their everyday lives. To now go into another platform to go through a whole other set of workflow is just something that they dread regardless of how good it is for the patient or the economics. It's just more work.

So getting our system to be easier, faster, simpler, more integrated into their daily workflow and to use the power of AI to where -- so we take their dictation. Every doctor dictates. So if we can take their dictation and then take -- and then pull apart the data in that dictation and populate the forms for them, those 2 or 3 or 4 minutes that they save is massive to them as they -- and that multiplies with each patient that they see.

So we are very committed to making our workflow now. We've solved the major problem of connecting them and aggregating data. And now we're going to be incessant and passionate about making it easier and easier and easier and easier for them to implement, for them to use on a daily basis, for it to be intuitive and to let the hospitals continue to have the type of ROI they need of capturing all these images. So -- and we're never going to stop making it, and we're going to use every new tool available to us in AI to make it easier for our customers.

Our next question is from Ben Haynor from Lake Street Capital Markets.

First for me on the -- so just on your own internal development pipeline, P5.1, when should folks expect you to go to the FDA with that? Anything new on kind of the new form factors, whether that's iQ station or some of the wearable sort of form factors that you showed at the Investor Day a couple of years back?

Yes. So thank you for the question. We're hoping in the beginning of '27 to have the 5.1 chip into our next probe. So that's kind of our timing. We think that -- we just sent it into production. Normally, it would take us sometime late this summer to get them back and then we start putting it into the hardware, validating, testing and then hopefully get our clearance and then we'll be out sometime early of '27. So that's kind of our schedule.

And at the same time, we have a vision to bring one-to-one in hospitals and how we can make a cart-based experience with Butterfly better. So iQ station is actively on our road map. And it's probably a later 2027 time frame for that. But we are actively working on that program. And we think that will really codify the one-to-one model on the enterprise side.

So we're working on making software easier. We're working to make our devices more powerful. We're working to make the workflow in the hospital easier. So it's just a -- it's just inevitable what will happen, and we're on the right path there.

And then on the wearable side, if we wanted to launch a wearable tomorrow, we could or theoretically, we could, we can put that. Wearables is not an R&D thing. It's more of a what's the use case and what's the market and what's the business. And so our philosophy right now is the success of HomeCare will determine what our wearable use case is.

And the more that we push ultrasound outside and we push it into nursing homes, we push it into the home, and then the more we'll be, okay, now this is what it's being used for. This really makes sense. This is sticking. And now let's automate it and let's make it easier and let's put a wearable out there.

So we have -- there is one of our partners, Forest Neurotech took our technology and created a wearable. Go on their website, you'll actually see it. They put it right there. And so the technology around creating a wearable is not the challenge. It's, hey, what's the business use case? When does it make sense to invest the next several million dollars in hardware to put that out there? And so our HomeCare business, in my view, will determine the success and the timing of when is the -- what is the right use case for wearables.

Got it. That's very helpful. And then any more you can share on that POCUS innovation forum that you held? That sounds like a quite interesting event.

You probably need to come next year. We probably have to invite our analysts. We had it private the first year. The second year, we brought our whole management team and commercial team in to listen, which was a great education session. I think what's fascinating about that forum is it's not all one specialty. You have a cardiologist sitting next to an anesthesiologist sitting next to an ER doctor sitting next to a primary care doctor, sitting next to an intensivist, sitting next to a nurse and sitting next to an administrator.

And they all have a tremendous passion for the power of being able to do a diagnosis immediately at the point of care. And we all heard them present their experiences. But we also gave them a forum to tell us what they need. And that's where actually Compass AI came from the last year's forum, was when they said it just takes too long to document each one of these records.

And we went now from 4 to 5 minutes down to 30 seconds. And so we listen to them, and that's why this forum is so important. And I think while we had 60 people -- we had about 40 people in the first year, 60 people next year, we'll probably open it up to 100 or more and even maybe allow you guys to come too.

So we learned a lot. And the biggest thing that we learned is that we're on the right track. And if we keep listening to our customers and being honest with ourselves on what we need to do to improve, we'll win. And it was a great session, Ben.

Sounds fantastic. And lastly for me, just with Midjourney, you phrased it as a revenue share potential rather than a royalty. Should folks read into that in any way? I mean that strikes me more as a service offering rather than a device offering.

No. It's -- I think you just look -- you look into this like a partnership. We are all in with them, and we are doing everything we can to make them successful and we'll do everything we can to make them successful. So we don't look at them as someone we're going to sell chips to. We look at them as someone that we are going to invest in, and we are going to do everything to amplify their use case. And if all of a sudden, we need to do something that was not planned, we're just going to do it because we both believe in what they're doing, and we're going to be the best partner possible.

We currently have no further questions. So I will hand back to Joe for closing remarks.

Well, thank you all for being with us this morning. I know that this time of the season is busy for all our investors and analysts. I hope you can see the excitement. I think the most important thing for me is we put a plan together, and we are executing to our plan. And everything that you hear on this call, there's not one surprise, not one.

Everything that we are delivering, we've told you we're going to deliver. Everything that we are focusing on are things that we've mentioned, if not recently, a year or 2 ago. We are executing to our plan because there's a massive opportunity in digital ultrasound. We are digital to film, and we are digital that will take over analog. It's just a fait accompli.

We've put a plan in place. We are executing to it. We have an awesome team. We are beyond the time where there was doubt about the company and the only doubt should be in that -- for anyone who's in our way.

So I appreciate your attention and look forward to continuing to deliver results. Thank you.

Thank you, everyone. This concludes today's Butterfly Network, Inc. Q4 and FY 2025 Earnings Call. Thank you for joining. You may now disconnect your lines.

Hi, everyone. Thank you so much for joining us today. My name is [ Lina Campbell ] and I am an associate with the health care investment banking team at JPMorgan. It's my great pleasure to introduce Joe DeVivo, the CEO, President and Chairman; and John Doherty, EVP, Chief Financial Officer; and Megan Carlson, the Chief Accounting Officer of the Butterfly Network team. Please join me welcome the team here.

All right. Well, first of all, thank you so much for the invitation. It's a great honor for Butterfly to be able to present at this JPMorgan Conference, the Super Bowl of health care each year. And thank you for all of you being with us on this last day. So we're very honored.

But before I get into the presentation, this morning, there was some news, a press release, where we've had at the SPAC, we issued a bunch of warrants to select shareholders that had helped the company with the SPAC at $11.50, and those warrants are expiring. We knew this would happen. We would -- of course, we wish our stock right now was at $11.50, $12.15. We think actually it should be there. But with those warrants expiring, there was some news. And so I think people were confused when they saw that press release.

So we knew this was going to happen. That actually means it's unfortunate for those warrant holders, but for existing shareholders, new shareholders that's 20 million shares of dilution that won't be happening in the next month. So no big deal to us, and again, except for the warrant holders who've actually been very faithful to the company. We're sorry for that, but we'll get it up there, and we'll get the stock running, which it has.

Butterfly is such an exciting opportunity. I have a lot to share with you today, so I'm just going to get right into it. And we will have forward-looking statements as a public company, so please be aware. So just a little bit of a history lesson. We started this company in 2011, the first semiconductor on ultrasound -- ultrasound on semiconductor was developed in 2015, and that was our first version of Poseidon, our Poseidon family.

I will talk about our family a little bit later. But that first one was 2015. That's how long we've been working at this. We've launched several probes since then. We've been following Moore's Law and as our semiconductors get more powerful, we launch new probes with greater and greater imaging. We did go public in '21, and we now have launched Compass AI, an incredible software that I'm going to detail for you in a moment.

What that does for us is that in ultrasound, there are different frequencies, there are different arrays for different parts of the body. And semiconductors used to be huge or supercomputers, I'm sorry, used to be huge. Now you all carry one in your pocket. Ultrasound devices are huge. But every doctor and every nurse in the world in the future will have access to be able to have an ultrasound device in their pocket.

We are the stethoscope salespeople 40 years ago when they were convincing people to look inside -- or listen to the heart sounds inside the body when you're with a patient. We are now convincing doctors and nurses that they can look inside the body when they're with a patient to try to find earlier diagnosis, to be able to find things sooner. And then if they do, they confirm it with radiology. This is not in conflict with radiology. This amplifies radiology by increasing the funnel at the front end.

But having one device that can do everything is so powerful. We have 1,000 of these in Ukraine. Every medic -- most of the medics in Ukraine have these when they encounter a soldier. We have these in conflict zones. We have these in Africa, in rural countries where 2/3 of the world doesn't have access to medical imaging. This is 1/10 the cost of an ultrasound cart. It's AI-driven. It's cloud-driven. It's semiconductor-driven. It's just awesome.

And so we've been building a brand. We have the largest installed base in the world. People understand what Butterfly are -- Butterfly is. About 80% of the medical schools teach kids on Butterfly. Kids -- these medical schools want to learn ultrasound, they want to have this as a core competency. And it's not lost on our top culture. This is the #1 TV show in the United States right now and just watch.

[Presentation]

Did you butterfly the patient? Grab the butterfly. Let's butterfly that patient. That's developing. When your brand starts to become a verb, you're doing something right. And we did not pay for this placement. They called us and said, can we have a new probe? How does it work? And we taught them the medical -- someone on the medical team is a big Butterfly fan. And the pit is famous for asking emergency room doctors what they go through, what they experience. Tell me about a case or a situation when you had to deal with X,Y and Z, and those are those situations that you're seeing on TV. It's why it's #1 show because it's authentic, and it's representative of what's happening in health care. And it's also why Butterfly is so awesome.

And we can do virtually all these different specialties. You simply have a butterfly, you plug it in a phone or a tablet. You say, I want to do a cardiac, I want to do pulmonary, I want to do a carotid, I want to be able to do MSK. And the device because the semiconductor on chip or the ultrasound on chip, it changes the settings, the second you say what you want to do and then you go. That's why in the emergency room, you don't know where you're going to encounter. And so you have the ability to toggle it. But if you're a specialist with a certain part of care, that's great, and you want to focus on cardiac, that's great, but people also want to have the opportunity to look elsewhere in the body.

We are an award-winning company, not just popular on TV, but the Prix Galien Award was the #1. It's like the Oscars of the best medical technology. And we're also, at least to date, as far as I know, unless anything else has been announced, we're the only medical technology company to have an Apple Design Award.

So for those of you that follow butterfly, as butterfly gets more and more popular, doctors, of course, use it in their practices. They use it outside the hospital. But when -- now there are doctors where there's 50 -- there's 100 doctors in a health system that all have their own butterflies. And when they go into a health system, that data is not captured, and that's not a good thing. We should have that record in the patient data, the images should be in the DICOM and they should also be filing for reimbursement. About 35% of all the point-of-care ultrasound scans that are done in hospitals actually make it all the way to being reimbursed.

So they're leaving about 65% of those scans on the table, which is those ghost scans are dead revenue. And so we've been selling -- we launched this a couple of years ago, a Compass software. We've been selling it to hundreds of health systems throughout the United States and in different departments. And then several health systems where it's enterprise and in every department, but we've recently now made a major update. And that major update is by making it easier because while hospitals like capturing those additional images, doctors are like, ah, this is another thing I got to do. This is another step I have to take. So with all these advanced AI tools, we wanted to make this easier for them.

So they have to fill out like a document with 15, 20 different fields in order for that document to be complete. I'll go back and I'll get us started properly. Sorry guys. So instead of them having to fill out each of those fields, what does a doctor do after a consult. They do a dictation, right? So right now, this doctor is doing a dictation. And our AI tool is listening to that dictation and, okay, great. So now we have speech to text.

So in that text are all the individual data points that you would need to fill out the document. So instead of the doctor now having to go and fill out the document, our AI tool goes through the dictation, pulls out each of the data and, bam, populates it all into that file, pretty cool. I mean it's something that allows for the doctors to win. So they get the document, but they don't have to sit there and toil through it.

So what they do then is make sure it's accurate, it's complete. The file won't go through unless all fields are there, and so it will be in red, they fill it out, and then they're done. So we are continuing to innovate. And when we talk about we believe that every doctor that's applicable for ultrasound should have their own probe, just like every doctor has a stethoscope. They should be able to listen inside the body. They should be able to see inside the body. And because there are diagnoses and things that happen that if you have to wait months for or you miss it, disease doesn't wait. It progresses and capturing them sooner is the essence.

And so Butterfly is the company that has the greatest scale in all of point-of-care ultrasound because we have the software, we have the devices, we're cloud connected. And all of this works in a seamless ecosystem into the EMR, into the DICOM and into revenue cycle management. So when we now -- as we mature and hospitals start saying instead of all these doctors just buying them on their own, we're going to now buy them and give them to everybody. Butterfly is that enterprise company that will win.

And so what we've done now is we started to open up our platform, because we realize that there are a lot of other capabilities that can come through butterfly. We have the largest installed base of point-of-care devices in the world. And there's a lot of people who would want to access them. So we've started to compartmentalize our technology and then open our technology to third parties.

Now let me just -- let's -- there's a lot of young people here, but some of you are old enough to remember this. Who here has ever had a BlackBerry? Yes. All right. Well, where are BlackBerries today? We initially had to have a BlackBerry and our institutions told us that those BlackBerries were the institutional lockdown device, this is what you have to use. What killed the BlackBerry? Not just the iPhone, apps. All of a sudden, every day, another app would come out into the marketplace, another application. And now wait a minute, you mean I can do GPS on my iPhone. I can get to TomTom off my dashboard, and I can now follow myself on the phone with a GPS app.

So we have 27 companies today that are developing AI tools that will link into our software development kit that our installed base can purchase, put on their phone, pay them for that license and now that application, whether it's a cardiac echo or whether it's a pulmonary congestion scan or whether it's for deep vein thrombosis, a lower-level clinician who's not classically trained in ultrasound can open an app and an AI tool can help them capture an image, a nurse in a home or a doctor in their practice who wasn't classically trained. While 80% of medical students today are being trained on Butterfly guys.

So this is done. The die is cast. We've already won the game, right? But it's probably that 15- to 20-year victory waiting for all these kids who are in school now to be out in medical practice and changing the market. So how do we compress that sooner? We make it easier because ultrasound is the hardest technology to do, read and capture than all the others because they're much more automated. But with AI tools, it's going to explode. So we are very excited about this part. We've opened this as we called Butterfly Garden. And it's happening.

The first FDA-approved app came out last quarter and that was HeartFocus from a company called DESKi. And if you were a lower level nurse, you can just simply grab the probe, download the app, place the probe on the chest and then the device starts -- the app starts telling you, wait a minute, move it here for a better image, twist it, push down, go enter an angle. And the quality score gets higher and then, bam, it takes a picture, and it walks you through that protocol.

So now a nurse in the home can take a cardiac echo, which is a $500 scan that you have to wait months for in the hospital. They can take the echo, they can send it to the cardiology clinic and they can make a diagnosis. This is changing health care major velocity, and it's happening. And these are some other education apps that are available today. But now every quarter, more and more apps are going to come out and those BlackBerries in the hospital are going to be obsolete.

And we're also not only devices, not only platform, but we're also service it. And so we are working on home care and coming up with new models that can -- that will take the risk. We will work to educate those nurses. We will use these AI tools, and we'll show at-risk providers how to be able to care for these patients at an accelerated pace. So we're really, really excited about where we are.

But guys, something else is happening out there in the world. AI, we all know the explosion of AI. But there's also some people out there with some fantasies. And those fantasies have to do with a convergence of AI into the human body. And so as AI wants to communicate to the human body, there has to be a portal. There has to be a language. There has to be something digital that can communicate. Now we've heard of brain-computer interface devices from Neuralink where they -- and other companies where they place wires in the brain. Well, those wires -- the biggest issue with those wires is something called the butcher factor. How many neurons do I have to kill before I can get to the neuron I want to talk to. With ultrasound, we do something very unique. We have a chip and that chip can actually move and steer the beam. So if you implanted it into the brain, that chip can scan the brain and see 25% of it just with one chip. And so other companies are starting to come to us and say, you've spent $300 million developing this chip. You have 600 patents. We want to use your chip inside of our product.

And in the past, we kind of pushed these companies away. We had some non-health care companies that came to us and we were like, oh, well, we're too busy building point-of-care ultrasound. Well, guys, I think we have a whole new business opportunity for this company. There have been companies in the past that have solved major technical problems for themselves. And then when they allowed other people to use it, built massive businesses that they never intended to build. Anyone remember a company called Amazon. Do you think that they built -- went out there and built cloud computing to be a cloud computing company. They built cloud computing because they didn't think anyone else could do it at the scale they needed. And after they spent hundreds of millions of dollars in data centers, they were at 1% capacity. And they said, let's go sell some of this capacity. And today, AWS is the biggest profit or bigger profit contributor for Amazon. Sometimes, you solve problems that everyone else has.

We have a unique semiconductor that has 9,000 sensors on it, each sensor like a pixel on a TV screen. Each sensor can be operated independently. So you change the color of that pixel and all -- and then you change the color of the next pixel, you change the color of next pixel, you sit back and you have a picture. Well, what if we allow people to be able to use this to program for their own methods of communicating for whatever product they have. It's a massive opportunity.

We use this for ultrasound for our purposes. We can do all the different arrays. We can also do things that the competitive devices or the legacy analog devices can't. Ultrasound legacy uses a piezoelectric crystal, that crystal is a cut -- physically cut crystal. Imagine having a camera that's tuned for nearsighted to see close, right? No matter what I do with the software behind that lens, that image, that beam is going through a lens that's going to only allow it to see near. So I got to go pick up another lens, and then I got to go pick up another lens. That's why legacy ultrasound 40 years old, you have to have all these crystals. And companies who say that they have all-in-one devices now, just pack it with a bunch of toxic lead.

So we have this infinitely programmable chip that we've solved a foundational problem and the programming of it is very exciting. And so what's happening now? We're all aware of cloud compute, but compute is getting now closer to those who want the intelligence. So now we're hearing about edge compute for AI. We're talking about bringing AI compute nearer to the use case, and that is the next wave of AI. And so we're seeing devices in the home where there'll be some type of computer in the home with the AI and everything will compute off of that. So it doesn't have to go to the cloud, have the latency, get out of a network, et cetera. It can compute locally. And that actually has benefit for health care.

So as we have this powerful cloud compute that's creating all this value, the next wave is now having compute closer to the patient. And a lot of medical institutions like that because they don't want to traverse a firewall, they don't want to have latency, they want to make a quick decision. And so in health care, edge AI compute and AI compute actually on the device is going to be a very, very big market. And people believe AI and computer vision in its own is going to see billions and billions of additional market growth and market value.

So today, Butterfly is a point-of-care ultrasound company. I showed you this slide. We are in all these medical departments. We are in all these different locations around the world, but we are now opening up our platform further. And that POCUS is now only one part of a much larger TAM that our company is going to be able to enjoy because our chips can be used for patient monitoring, pharmaceutical applications, bone healing, endoluminal applications, HIFU catheter-based IVIS, robotics, neuromodulation, tomography. And we are having conversations now with all these companies who want to come in and license our chip. Pretty cool. Pretty interesting.

And just like the other companies who solve those foundational problems, it creates an entirely new revenue stream, an entirely new investment thesis for the company. And we've been proving this out. So our point-of-care handheld market opportunity itself is -- from analysts is around a $2 billion opportunity. Hospital ultrasounds about $6.5 billion and then getting ultrasound into long-term care facilities, we believe is about $1 billion where we can actually -- the ultrasound device of the future is going to be this size. This is my AirPod case.

They're going to send the devices home, put it on the patient, press a button. It's going to do a scan, send it to the cardiac nurse, all digital, all based on semiconductors. But now with that chart I just showed you, these are all the additional market opportunities that Butterfly never had before. So if someone wants to come into our data center theoretically for that analogy, we now are going to be able to sell that capacity or we sell our chips to noncompetitive companies who want to license our technology and who want to benefit from 600 patents, 15 years' worth of work, the best chip fab in the world, TSMC, behind us and a world full of knowledge.

So now this -- many of you have heard of Octiv. We talked about Octiv. We initially talked about Powered by Butterfly and then it became Octiv because we thought we would spin this business out. We even thought we would sell a part of it to get some nondilutive financing into the company when we were a little bit more in need of capital than we were today.

So we are not going to spin it out. We are not going to have it as a separate company. We're going to become that company and our -- and this program will be branded with our customers Butterfly embedded because it's software, it's devices and it's our chip.

So our Poseidon family, we, today, are on iQ3, which is our P4.3-chip, and it's been amazingly accepted in the marketplace recently. We just sent to production our 5.1 chip, which would be the last chip from the Poseidon family, which doubles the mechanical pressure and improves harmonics. And what you can see in this image is harmonics, when you have higher pressure, the sound wave flows through the body better and gets a better image.

Every big ultrasound company walked away from semiconductors because they said they will never get to harmonics. Well, you are wrong. Now we are -- now that our 5.1 chip is sent into production and we -- and that will be out in the next platform in the beginning of '27, we've started developing on an entirely new family of semiconductors that will have 20x the processing power of the current chips that we have today. And we have a road map for Apollo 1, Apollo 2, Apollo 3 and Apollo 4 and something that we'll introduce at some point in time as the super chip. But this chip will have the capability of not just edge AI, but running AI on the chip, on the device, having complex algorithms, providing value to clinicians.

So as I mentioned, AWS, there's another analogy. I don't think we'll ever be as big as NVIDIA. But you got to remember, from 1993 to 2005, NVIDIA was a video game company guys. They made CPUs and GPUs. And they focused on what's the best gaming they can do. And in 2006, they launched a developer software called CUDA. And that developer software allowed them to open the platform to that third-party development. These are things that have been done before. And these are things that now we think we can do again. We think this market opportunity is huge, and we have already partnered with up to 3 different brain-computer interface companies, one of them that we've talked about publicly, Forest Neurotech; a vascular robotics company, Mendaera; a fatty liver company, called Sonic Incytes. And many of you who've been looking -- searching the case have seen this announcement we made with Midjourney, which is a GenAI company, which is a text-to-image GenAI company, who within the terms of the agreement, this is about a $74 million deal that we signed with them for them to have exclusive license to our chip to a certain specific domain.

I can't tell you what it's for because that's their opportunity to communicate this publicly, but it's a big wonderful opportunity, consistent with everything I just mentioned. They're an incredibly innovative company and I think you guys will be thrilled when they choose to unveil it.

Last thing I'll say, and I know I'm a minute or two over. I'm sorry about that. But Butterfly has been through a lot. I came here in 2023, when our revenues were declining and we were consuming hundreds of millions of dollars of cash. Within the last 24 months, we have accelerated our top line growth. We have improved our gross margins and we've dramatically reduced our requirement to consume cash.

We've raised money and are now even within these licensing deals, bringing in more nondiluted capital now into the business. We are completely poised and completely strong on our own balance sheet. We are executing as a company and it's just going to be a great time forward. So I just wanted to thank you for your time and attention today. This is a great time for Butterfly that the restructuring, the questions are all in the past. It's now go go go. Thank you all.

Thank you very much, Joe. First, I wanted to say huge congratulations to you and your team for all the achievements to date, definitely very impressive technology and development.

Thank you.

I do have some questions for you from the audience here. The first question we have is, as we look ahead at your next generation of semiconductor chip, what should the stakeholders expect in term of performance, capabilities and strategic focus?

So the next chip is our Poseidon 5.1. The 5.1 delivers double the mechanical pressure of all the other chips that we put together in the past. That mechanical pressure by increasing the voltage through our MEMS has allowed us to get to something called harmonics. And I like -- my analogy for harmonics, which would make an engineer cringe is if any of you have been on a boat and you just hit the accelerator, the first thing that happens is the bow goes up, right? And then you just wait for a little bit and then it starts to glide and then it planes over the water and it flows with less and less effort.

When you increase the mechanical pressure at that site, it actually allows a sound wave to flow easier through tissue. And when that happens, the image quality is dramatically higher. This is the big miscalculation of big ultrasound why they didn't take a chip to market and why our founder was so visionary to not only get into this but believe that we would accomplish this goal.

Now that we're at harmonics, everything is about processing power, and you saw what will happen with our next generation. But this is now in production over at TSMC. We will then put it into a device at the end of -- middle to the end of '26 and have something in '27. Thank you.

That's super insightful. I do have more questions for you here, but also wanted to make sure we take questions if there's any from the live audience. If you have any questions, please let us know, and we'll have a mic pass around to you.

Cool. I actually have more for you. So could you tell us more about the Apollo chip. What are the key benefit it brings in areas like compute, efficiency or AI enablement?

Thank you. So Apollo is a completely different architecture from the initial Poseidon family. There are some things we were able to take out of the chip, which created that can be done on other components in the handheld, which allows us to have much more room for an additional level of compute. So that -- and that's a 20x compute on top of that, which will translate into the ability to run AI locally but also the ability to continue to march up our image quality.

We believe there's opportunities to make ultrasound look easier and get to 3D and to do other things, but you have to have that compute power at the place. A lot of the -- the reason why these ultrasound machines are big is because they're taking analog signaling. And then they have to go through all this compute to convert analog into digital and then they represent digital on a screen.

We don't have to go through that compute because we're acquiring it in a digital format. The other thing about Apollo is it will use a lot less energy. It's much "cooler". And so by using less energy, it means it's more and more amenable for implantation. It's more amenable for longer battery life for smaller form factor. We -- the gigahertz in a -- like when Kodak launched their first digital camera, it was only 1 megabyte. There's only 1 -- and the pixelation was not as great. And so they jettison the program. But once they got to 3, 5, 7 megapixels in 2003, next thing you know the image quality was as good as film and digital took over analog. Well, everything has to do with the ability of having those increased megabytes. And now you look at the iPhone not only did they launch a 48 megapixel image recently, but the camera is only taking about 10% of the real estate of the phone.

So everything -- and Apollo is a major leap to not only increase compute but have the ability to be smaller and to take less resources. And so it's completely the next and the right chip for the next wave of AI and the next wave of not only ultrasound, but a lot of the things that we talked about that third parties very much are looking for.

We also have a question here for you related to the POCUS business. Could you tell us more about what should we expect in your POCUS business in the near-term strategy? And looking forward, how do you see chip advance influencing the future of the core business?

Absolutely. So thank you so much. So we've put out about 150,000 units since our inception and there are 27 million doctors and nurses around the world. So that's a massive market opportunity. We haven't even gotten started yet. And what's wonderful about our product line is that we've kept our second-generation product on the market, which is still very viable, and now we have our third generation. Our second generation is priced close to just a bit above $2,000. So that device has a lot of popularity in our global health initiatives.

We partner with over 1,000 now NGOs who do medical missions and all different type of care for people around the world. And so we have the lowest priced device in the world. And we're also moving up the curve. So our next-generation iQ3 is $38.99. So it's more expensive. It has more processing, but we kept both on the market because we have the lowest price, and then we have the next version. And I think we're going to continue to look at what our portfolio looks like across the board.

Probably one of the most important numbers on the slide, I didn't talk about, was we have preannounced our fourth quarter a few days ago. And we grew 17% year-over-year, and that's off of a quarter that grew 35% the year before. So our core business is well. We started off a little slow because of some macroeconomic things that held some of our biggest projects down. But we had told the Street that our fourth quarter was going to be a lot better and that we saw things coming back, and that number indicates it did.

So we're very bullish about our point-of-care ultrasound business. Just like I said before, there's going to be a killer app out there. One of these companies with one of these AI tools is going to be something that everyone absolutely has to have. And when that happens, we're ready. We have an incredibly scalable supply chain. We can go -- we sell between 25,000 and 30,000 devices a year. We can go to 100,000, 200,000, 300,000 like that. So when the market is ready, we're ready.

We're also all cloud connected. We have 25 million images in the cloud, and those grow 30,000 to 40,000 new images a day. So what's the most difficult part of developing AI? It's not the algorithm, it's having the data to run the models on. And we have the largest private database of medical ultrasound images in the world. And so we are the AI developer for ultrasound, and we have all of our partnerships, and we're developing our own AI tools as well.

We are the device company that has the ultrasound chip with the different form factors. We are actively looking to develop our home care business and ultimately, a wearable device of a size like this that patients have. And we spent the first part of our evolution as a company trying to become equivalent to everybody else, which means you might think you're inferior. Oh, our image quality is not good enough.

Well, yes, like in the beginning, when you're starting off with the semiconductors, it's a lot to get out in that first version and first generation. But now as we are marching up Moore's Law, we have put more new innovation out in the marketplace than on the transducer side than 40 years of legacy analog ultrasound. So we are no longer worried about being as good as everybody else.

Our focus as a company is we're going to change the world. We're going to change the way images are acquired. We're going to change where they're acquired. We're going to change who is acquiring them. We're going to increase access to care for patients who need it. 2/3 of the world don't have access to medical imaging, they will now with Butterfly.

We will keep our prices of those generations low. We will work with NGOs. We will work with governments in need. We have an app right now in 2 African countries that you can -- for many pregnant women don't -- in Africa don't know the exact date of their conception. And if you don't know your date of your conception, as you look at each of the phases of a woman going through having a child, they need different types of care at different stages. And if you don't know the age of your child, how do you know you can be given the best care.

So there's an effort out of university in North Carolina. We built the app directly into our device, and we've now launched it into 2 African countries, where you can simply take the probe over the pregnant uterus and scan it, and it will tell you the age of the child.

Is that insane? Yes, it's insane. That will change the world. Butterfly is in the process of changing the world and having a mechanism to deliver this intelligence in a small, cost-effective, durable, mobile way is the gateway to all of this intelligence getting to everyone in need in the world.

It's a very strong momentum with the business. I know we're close on time, but I'm going to squeeze in one more question. We have one more question here related to your codevelopment opportunities. Could you tell us more about how do your future generation chip shape the evolution of codevelopment opportunities, the new possibility unlock from a partnership standpoint with Butterfly embedded?

100%. So what's happening now is we're having conversations with people who want to use the chips, but they're using [ workforce ] for different applications. And one of the things we do when we get the material from TSMC is we can get it in a wafer form. So instead of just having the chip come in it's cut predetermined form, we can get the wafer and we can cut that wafer then in different sizes if you want a small one for a catheter, you want to be able to have something very small to put on a lens, you want something round, oval, et cetera, et cetera. So we are now working with these partners to identify how to meet their needs, how do we take our core technology, our core hardware. And it's not just the hardware, it's the hardware, it's the MEMS, it's the AI, it's the software. It's everything that we have developed in order to make this platform amplify.

And so when we meet with our partners, we create the road map and we develop with them. And so far, the partners pay for that development. So if the partner pays for development, that's not on our road map, that's incremental to Butterfly, and that's also incremental knowledge and additional capabilities that Butterfly brings in. This is a new era for the business. I cannot be any more excited and proud of my team. I can't be any more excited about the technology, the executives I have with me, Steve Cashman, in the back room who's driving our commercial business and our vision. We have an unbelievable opportunity here, guys. And you just keep watching and pay attention. Thank you.

Thank you, Joe.

Thank you all for joining. I'm Daniel Markowitz. I cover life science tools and diagnostics and med tech here at Evercore ISI. I'm joined today by Butterfly Network. With us, we have Joseph DeVivo, President, CEO and Chairman of the Board; and we have CFO, Megan Carlson as well. Thank you guys for joining.

Great. So I wanted to start off pretty high level. If you wanted to just give an introduction to the company, what does Butterfly do? How would you describe sort of the market you play in? That would be a helpful place to start.

So medical imaging is normally historically been something that's [Audio Gap] over the last 15 years, especially starting in emergency rooms, doctors started to say, I can be better if I had the opportunity to do the ultrasound myself. I can do an immediate, especially in a trauma setting instead of having to stabilize the patient, send them to radiology or order a scan and wait either for someone to come down or to send someone up. So that initial thought kind of created this idea of, well, why doesn't everyone do ultrasound at the bedside.

Now ultrasound machines are big. They have multiple probes on them. Each probe has a different frequency and array for different parts of the body. So it's -- there's only 82,000 people in the United States today who are classically trained in how to do an ultrasound, and they're called ultrasonographers.

So over the last 15 years, doctors have started to say, like -- and especially a Hennepin County up in Minnesota, the emergency room doctors said, hey, instead of buying the next -- well, when you buy the next ultrasound machine, give us your old one. And they started to scan people and learn certain cardiac scans and find that they can dramatically improve the course of care for those patients by simply looking and seeing.

And so over the last 15 years, we've now seen emergency rooms, pediatric clinics, primary care, anesthesiology, intensive care, cardiology, OB/GYN, all different types of specialties build their own capabilities to be able to do this type of imaging right then in there as an earlier diagnostic step before just sending them to radiology. And especially when it's emergent, you can intervene care. You can see things sooner.

So companies started to cut the cord on each of those probes that are on the machine, put a battery on them and sell them as a handheld ultrasound device. And also ultrasound companies I've started to make carts that aren't catch-all type of carts, but ones that have specific use cases.

But then there's Butterfly. What Butterfly did was each of those devices that are currently in the market use something called a piezoelectric crystal. It's like a lead crystal that's precut to deliver a certain array. And think of that as a lens on a camera that's not auto focus. So it's like I can see far, I can see near I put a new lens on or I have to pick up a camera with a different lens. And so you would need multiple devices with -- that can look at different parts of the body.

So Butterfly came along and invented the first ultrasound on a chip. So it uses a semiconductor and then has a wafer that goes on top of the semiconductor with 9,000 little drums in the rectangle. Think of it like pixels on a TV screen. We can control each one of those drums and program to deliver any type of array of sound that you'd want to do. So the benefit of that is instead of having multiple devices for different use cases, you now can see the whole body with one device. So our vision is that every doctor and every nurse in the world will have a Butterfly. 27 million doctors and nurses around the world. and everyone will have that. And over the last 10 years, we've been developing the semiconductor with more and more sophistication.

The first device that came out, the image quality was not at a cart level or at a level where people felt in subspecialty care, they can make those clinical decisions. But Butterfly is on this rapid innovation path of Moore's Law and compute power improving dramatically. So we are now on our third-generation chip, which has our image quality, we believe, as good as any other handheld on the market. And instead of having to have multiple handhelds, we are completely cloud-based. We're completely mobile. We plug into a phone or a tablet. And 2/3 of the world today doesn't have access to medical imaging. But now Butterfly is military grade and is used in third world countries. There's 1,000 of them in Ukraine, where medics -- trauma medics are using them to stage and triage people who have been harmed in the war.

So we're providing a massive service in a massive market to democratize imaging, lower its cost. Our one all-in device is less expensive than the individual multiple devices you buy from other companies. And we're really leading the way.

Got it. That's really helpful background. So if you could talk about the goal in the future, every doctor has one. I guess what's the current paradigm? I guess what percent of ultrasound images today are taken on traditional machines versus handhelds? And do you see that niche being in those markets where you can't do the medical imaging, places like Ukraine? Is it more like niche applications? Or where do you see the adoption kind of evolving with time?

The adoption is that this is the stethoscope of the future. We're the stethoscope salespeople 40 years ago. Hey, doc, it might be really beneficial to you if you can listen to the heart or if you can listen to bowel sounds or hear fluid in the lungs. You should carry this around your neck and have it with you all the time. So let's just replace the listening part with the seeing part. This is a stethoscope of the future. And it's not just -- and yes, we provide low-cost access for third countries, but a doctor today can walk into their medical practice, which many of them do, pull out their Butterfly and make and look inside just to see, is it -- let's look and see if you have fluid in your lungs. Instead of just ordering a chest x-ray costing the health care system $500, just pull this out. And actually, it's more specific and it's a better image than a chest x-ray. And then they can go to a clinic. Then if they see a patient in the hospital, they have their Butterfly with them like they have their stethoscope.

So no, we really think that every doctor in the developed nations and of course, there are countries if they can afford it, will have a Butterfly and every nurse also. Because what's also empowering this, there's a reason why there's only 82,000 ultrasonographers is because while ultrasound is the lowest cost imaging modality, it's the hardest to capture the image and the hardest to read. So it's like you could be in a dark cave. Now if you have a flashlight and you're looking for something specific, you have to know where it is, right? So you have to be able to locate it. And then once you find it, you need to know what are you looking at? So ultrasound is hard.

What's going to liberate this and allow this to get to the 27 million people around the world is AI. So we created the Butterfly Garden, and we are developing as well as 27 other companies are developing advanced AI tools to make it easier to identify something, easier to find it. And we have automated tools that you'll be able to place the ultrasound device on the body in the future, and it will just simply find something and tell you what it think that's there. And that is how we'll get past the learning curve and how we'll see mass adoption, just like for those of you who are old enough to remember when the iPhone came out, which I don't know if you are. You don't remember that?

Remember when you had a BlackBerry. Did you have a Black Berry?

Yes, of course.

Yes, everybody have a BlackBerry. Well, your institutions told you had to have a BlackBerry because that was secure and it was locked down. But what happened? A new app would come out. Oh, I can test my blood pressure with this. Oh, I don't need my [ TomTom ] on the dashboard of my car. I could do GPS with this thing. I can take the alarm clock off of my end table, and I can have all these different things. The iPhone has obsoleted hundreds of devices. And that's what Butterfly will do to the ultrasound industry because as every new app comes on, now I can do a DVT and I can open up a DVT app and I can scan for deep in thrombosis or I can scan for pulmonary or I can scan for cardiac echo or I can do fetal monitoring with it. As each of these apps come out, more -- it will -- the device will become more and more popular. People will carry it more and more than anything else, and then there will be a tipping point where it captures the entire market.

Great. Great. So what percent of the market today is in handhelds versus traditional? And I guess what's the friction to getting that number higher?

The friction, so from a specific number, we'd have to look at a market research report. I think the overall handheld market today is $300 million to $400 million. Now from a unit volume standpoint, I'd have to see that against like carts and all the other stuff. But ultrasound is a $10 billion global market today. So the actual revenue, but we're charging $4,000 a device and -- or, let's say, the average selling price is between $5,000 and $6,000 for our market, whereas the ultrasound carts go from $30,000 to $150,000. And so from a unit standpoint, I don't know what that market share would be. But obviously, from a revenue, you can figure out from the numbers I gave.

But what's happening though is as each new semiconductor comes out, like go back and think about digital photography. Now let's go back to photography. In '95, Kodak launched the first digital camera. They had 80% market share in film. They launched the first digital camera. Two years later, they shut the program. We shut it down because they only had 1 megapixel image. So it was a very pixelated image. It's almost like saying, "Hey, when I get Netflix versus going to blockbuster or I go blockbuster, I got a 4K movie, it works like that perfect, whereas my early Netflix buffered all the time". Whether it was my Internet that wasn't good enough or the software wasn't good enough, watching a movie that stops and then buffers like 5 or 10 times, you just -- no matter how good the idea is to stream, you're not going to do it. Just like in digital imaging, when the images weren't as good.

But the moment digital is equivalent to analog. It always wins because now I have -- I don't have to develop film, I can store it in the cloud, store it in forward. I don't have to develop images I don't want, I can modify them. All those same benefits are what Butterfly are doing to the classic ultrasound industry because they're using these lenses that can't be tuned and changed and improved. They can do things -- you can do things with software on the back end and push different signals, but those lenses, you can't change. And so for us, as each new semiconductor now comes out, what the film -- the photography industry learned was, as I got to 3 megapixels and 5 megapixels and 7 megapixels, all of a sudden, I got to image equivalency. And that's when the market in 2003 sold more digital cameras than film cameras, and they haven't looked back.

And now the cameras that we have are -- it's not -- I think Apple just launched a 48-megapixel camera the other day or a few months back. And it's not just about the image getting better and better, which it is, but it now only takes up about 10%, 15% of the real estate of your phone. So it gets smaller and smaller and smaller. So the imaging device of the future is going to be this big. This will be the ultrasound device of the future. And I'm holding up an AirPod case. And you'll just -- you give it to a patient, they'll just place it on the body, press a button, and it will scan inside to take the image and it will either send it to someone remotely or it will have the AI in it to be able to identify something and specifically pull out what that is. So that's the future of imaging, and that's where Butterfly is today.

So we just launched our third-generation technology. And we just did that a year ago. And now our fourth-generation chip is going into production. And that fourth generation will be out in -- at the end of '26, beginning of '27, and we're starting the development of our fifth generation chip. And the power and the ability to do edge AI is going to be in sync.

Yes. Got it. So pardon me if my quick math is off, but it seems like of the $300 million to $400 million market, you probably have about 1/4 of that within the handheld. So correct me if my math is off there. But I guess, can you just describe the competitive landscape? Are you mostly trying to go after business that's currently in traditional? Or are you trying to win from other handheld parts of the market? And then what do customers care most about? Are there certain specs that they're looking for? Is it data? Is it convenience? And then what like key areas are you focused on? You already touched on the pipeline and new innovations there.

So I think your math is pretty correct. The -- now if you -- there's a Signify market research report. And in that report, it has, for example, GE as a #1 in revenue by a few million, but they do about $15 million to $20 million in China, and we're not in China. So if you pull out China and you look at the markets that we compete and also, if you look at it from an ASP standpoint, from a unit volume standpoint, we're between 40% and 50% market share from a unit volume standpoint.

From a revenue standpoint, of course, we don't charge as much as all the other competitors do. So we give up on that part. And if you look at the U.S. alone, I think we're probably almost double our next competitor. So in the developed markets that have been really focused on this, we're pretty dominant in our space.

Got it. And I guess with the selling strategy, do you sell to hospitals, medical schools? How is the sales force structured? And is there an end market that's a greater level of focus right now?

Yes. We sell to all of them. And so it starts off all with online sales. So over 1,500 doctors a quarter will buy their first Butterfly online using a credit card. So we build a brand directly with the actual doctor. And so we have our online sales and then those who abandon their carts or don't complete the sale online, we have inside salespeople who call on them. We then have another inside sales team that calls on anyone who's not a hospital, so ambulance companies, large physician practices, clinics, et cetera, urgent care organizations. We have an inside sales team that calls on them. And then we have a direct sales team that knocks on the front door of the hospital walks in and tries to sell them not only probes, but our enterprise software.

And then through all of those channels, we sell the medical schools. Medical schools is probably where we're most dominant just because of our pricing, our ubiquity and we're in about 70% of medical schools today. So that's why this is a fate to complete. Like this is done. The die is cast.

The future of health care is every doctor and every nurse having their own ultrasound device. We're now graduating kids who've had their own device for 4 years, and they're going into their residency. So as all these kids come out and come in, they now have a skill set that everyone else or many others don't, which distinguishes them. And so we're just doing everything possible with AI and everything else to compress that time line to wait for the next generation but to do it now.

Got it. That's all really helpful. And then I know you guys have given a longer-term ambition for $500 million in revs by 2028, which you can run the math on what that implies. But I think there are some macro dynamics right now between global health funding, hospital budget constraints. Can you just talk to those and how I think that lands us at somewhere around 20% growth for this year?

Well, so we certainly expected going into the year that the hospital business would be a little more robust. But with some of the changes with -- first of all, with the NIH grants and then USAID, things just change for us. But I think that's over. But the USAID stuff is not, that's going to be more -- have to lean on more private donors. But I think the hospital side, it was a temporary budget kind of -- we're just not as important a sale in the hospitals and pretty easy to push us off. A lot of the big projects we had definitely got pushed. But I think that's going to be over. I think our fourth quarter will represent that. I think 2026 will be a better year. But I think our global health stuff -- it's definitely slower.

Got it. Great. And then just the last second, Megan, one for you. Just on profitability. Can you just lay out where we stand today versus a couple of years ago? And what's the path to profitability from here?

Sure. So I guess to start with cash. We ended Q3 with $148 million in cash after our capital raise earlier this year as well as we've been making a lot of progress on managing our expenses. We used about $45 million in cash last year. This year, we're expecting closer to $40 million or in the upper 30s. Our adjusted EBITDA loss for this year, we're expecting to be between $32 million and $35 million, which is an improvement over the prior year. So we're continuing to make strides. We're continuing to really analyze where we're spending and where we're investing. But that being said, we're not going to hesitate to invest in areas where we see an opportunity to accelerate some of the growth that Joe was talking about.

Got it. The last thing I'll mention is in order to get to that $500 million, we are stacking on top of our core business, new business opportunities. One of them is creating partnerships where other companies can leverage and use our chip. And so we just signed a deal, which is a $65 million deal over 5 years with a generative AI company that does text to imaging for them to build a new product using our semiconductor. And that's going to bring in more revenue that will be not only the licensing revenue, but the development revenue and then a revenue share. And we have a lot of other deals that we're working on in the pipeline. So for us to get there is not just an organic scaling. It's having the optionality of our core business growing, but then having our licensing business, which we call Octiv as well as a home care business and other objectives that should really get us into some hyper growth.

Great. With that, I think we're out of time. Joe, Megan, thank you so much.

Thank you. Appreciate it.

All right. Good afternoon, everyone. Thanks for joining us. My name is Danielle Antalffy. I am the U.S. medtech analyst here at UBS. Very honored to host to end the day, Butterfly Network. We've got CEO, Joe DeVivo. We've got Interim Chief Financial Officer, Megan Carlson. Guys, thanks so much for joining.

Thank you for having us.

Thanks for having us.

And it's been a few years since we've been following the Butterfly story. And I would love to get just your quick takes. I know we're going to probably get into more detail through the Q&A. But how the story has evolved, Joe? And you're -- what, you've been there for 2.5 years, 2 years.

Yes. 2.5 years.

Yes. Yes.

Well, I -- first of all, thank you for having us. And I'll apologize to anyone who's listening online or here. I have a cough. And if I cough into the microphone, it's going to be horrible. So I'll do the best I can not to. I remember within the first few months of joining the company, and I think I even said this on a conference call or something that I just felt like a kid in the candy store. When I had the opportunity to take the job, meet with R&D, I was just blown away with the depth of the technology that this company created from cloud to enterprise software to AI tools to an incredible semiconductor platform to the hardware. I was just blown away by the employees, the vision and the technical capability of this company. It's insane. And we're doing things today that no other large company is doing. And people, to a certain extent, still don't fully understand how capable this company is.

And so we've had this tough experience of having to resize the company in the face of that. We -- when I got there, we're burning -- or a Board member doesn't like that term burning. We were investing $160 million, $170 million a year. And that's rough. Like that's rough. And so we went through some resizing, refocusing. And then fortunately, we created a road map on how we would start deploying all this incredible technology. So what's different from when you were intimate with the story is, a, we've launched a probe that has exceptional image quality. When I got there, it was almost -- oh, Butterfly is so cool this and that, but the image is just not there yet. And it kind of likens to this whole time of digital photography when everyone is so excited about digital photography, but the image resolution wasn't there yet or everyone was so excited about Netflix but it's buffering, right? And it's like, well, I can't watch a movie because it's buffering, right?

But -- and so we were kind of at that stage. But now we've gotten to a stage where there is tremendous parity in handhelds now. So we have an all-in probe, lowest cost, does all these different use cases. And the image -- you don't have to sacrifice image anymore for that benefit. What we also did was we built an AI marketplace for -- there's about 50 different to 100 different companies out there developing AI tools to go into ultrasound. We then created a marketplace to allow them to link up to a public SDK and then develop. So we basically have gotten R&D -- so we have signed 27 deals. And each one of these entities now are investing to put AI into our environment. So that is new and exciting. And we haven't stopped our semiconductor development. So it's really, really exciting. We weathered the storm. We did a fundraise. I have to cough.

Go ahead.

Excuse me. We weathered the storm, resized the company, launched new technology, created new business opportunities, and we're in a great spot.

That's good to hear. So let me take some of the burden off your voice for a second here.

Thank you.

Yes. So -- and just so for folks on the line that might not know what Butterfly is, point-of-care ultrasound. Can you talk about where you sell into? Is it the health systems, the hospitals? Maybe talk a little bit about the business model for folks that might not be familiar.

Absolutely. So we make a handheld ultrasound device, and we believe just like we're the stethoscope salespeople 40 years ago. "Hey, doc, I think you might benefit if you can listen to the chest. You can hear the bowel sounds. You can hear fluid in the lungs and whatever." Now it's like, "Hey, doc, you benefit from seeing all those things." Now it's a lot easier to listen and to learn how to listen than to see, but it's the same thing. And we've made a device with a semiconductor that allows for all the different parts of the body to be seen with a single device. So there's a massive market opportunity. And the way we sell, it's something I've really enjoyed. I've been in health care for almost 35 years now and 30 years, 30 -- I don't know, I'll do the math later.

No one's counting.

No. Yes, I'd like to say 20 years or earlier, but I'm old. But what's really fun about this business model is we're digital first. So we have about 1,200 to 1,500 doctors every quarter buying our product online. So we invest a lot in social, a lot in digital, a lot in media. And that then creates the ability to build a brand directly with the doctor and a relationship directly with the nurse and the user. And if you look at how many people come to our website off of that activation, it's substantial. But the amount that actually gets all the way through the funnel to purchase is the number I gave you. So whoever doesn't purchase, all those leads roll down into an inside sales team that tries to activate them. And then we have another inside sales team that just focuses on ambulance companies, out of hospital, urgent care, and they do all that virtually. We just closed a major deal for a cruise line recently. So that was pretty cool to get them all on those boats.

That is cool.

And they did that all virtually. I'm so proud of them. They said, "Hey, do you want to go to the training on the boat?" I'm like, "No, I always told you guys go." So they're taking that trip.

Did they get to go on a cruise?

I'm hoping. They're supposed to. They're supposed to. So I'm very proud of them. They did a great job. And then we have a direct sales force that walks in the front door of the hospital.

Hospital. Yes. Okay.

And then internationally, we have distributors, and then we're direct in the U.K. and Germany.

Okay. And then it is the device itself, but isn't there a -- there's a software component as well. And if I recall correctly, you talked a lot about enterprise deployments, enterprise-wide. So maybe talk a little bit about that.

100%. So the way that works is there's a growing trend over the last 15 years of doctors being able to do ultrasound at the patient bed. So instead of ordering -- instead of sending the patient to radiology or waiting to have someone come to the patient bedside, who's an ultrasonographer, individual doctors have learned 1 or 2 scans and then built up their capabilities. And so it's called point-of-care ultrasound or POCUS. Now they do it a lot in the emergency room, they do it in intensive care units. They'll do it in anesthesia to place lines. But a lot of these scans you can get reimbursed for. But only on average, 30% of those scans actually make it into the EMR, make it into DICOM. So there's 70% leakage, revenue leakage for hospitals.

Yes.

So we built an enterprise software called Compass that connects to our device as well as our competitor devices and allows for them to program it where, first of all, they can collect all the image data from the devices. Second, they can actually document the encounter and that will then go into the EMR. The images go to DICOM and then when they're finished with the record, it pushes into the revenue cycle management software so they can get reimbursement. And so we've deployed these at scale. We have one health system with 65 different departments all on the software, and they've increased their capture rate to 77%.

Wow.

And that more than doubled their revenue, and it's huge. So we have this individual relationship with the doctors, but then we have this relationship with hospitals to help them operationalize point-of-care ultrasound, not just have the device, but then also the genesis of Compass was actually education because normally, when you have a new doctor, they're credentialing themselves on ultrasound, they need to be -- someone's got to look over their shoulder. So you can have a single administrator who's an ultrasound expert go into our software, go into that particular doctor's images, you can actually score and rate each image. And as they're going up their learning curve, you can graduate them up to certain places. And so that's what the software does also.

Okay. Got you. Yes. And so you had a win, I think, earlier this year. Is that right?

It is.

You closed a full enterprise deployment at a top 5 global health system. And you do talk about several large active opportunities. I imagine you can't give a ton of color on those until they actually close. But what differentiates the Butterfly product offering most often with RFPs today? So we touched on image quality. You feel like now the Butterfly focus is at parity with the other devices out there, TCO, Compass workflow, AI outcomes and evidence, like where does Butterfly win?

So we win anywhere someone is serious about deploying a program. And what's fascinating about this space is most administrators, even like we went to a conference in February, kind of like a boondoggle conference where all of these administrators go and then you can pay to go and meet with them. And so we met with 14 different Chief Medical Officers from health systems. They had no idea who Butterfly was.

Really?

They had no idea at what point-of-care -- one of them did actually, just one. Had no idea what point-of-care ultrasound was.

Interesting.

Because it's like a departmental thing where ER is like -- the ED is like, well, we need ultrasound, get us a machine, but it's not like they don't understand institutionally. So then one of the beautiful things about Butterfly is we are cloud-based, right? So we have 150,000 devices out there. I have 25 million images in my cloud. I have more images than any other nongovernmental entity in the world, and it's growing over 30,000 a day.

Yes.

And I can see every device. I know if they're taking an image. I know if it's not on or I know if it's -- if a scan is coming in, I know what type of preset. We have all this data. So when we sat with those 14 CMOs who didn't know who Butterfly was, we'd say to them, and like this one big customer that now has over 700 devices after they put the software in, we said to them, did you know 140 of your doctors have Butterflys? No. What? 140 of their doctors had -- they had purchased them on their own.

The doctors did.

The doctors did. It's like -- so...

But then you don't get the enterprise solution part -- component of it.

They're not getting any of the data, so they're not compliant.

Right.

And they're not getting any of the reimbursement.

Right.

And so when we sat with those 14 -- we knew who the CMOs were because they had to kind of select us to take the meeting. And when we sat with them, after they told us they didn't know who Butterfly was and they didn't understand what POCUS was, we would then pull up a document and say, 88 of your doctors have Butterflys. You've done 7,000 scans. It's mostly in this specialty, this, this and this. They were just like, are you kidding me? So we fight this major awareness. The big imaging companies focus on MRI, CT and then they'll bundle in the point-of-care. And point-of care ultrasound to a big imaging company is a kind of single type of use case cart. So like they sell these $15,000, $20,000, $30,000 carts that are more compact and that go to the ED or whatnot.

And that's the lion's share of their focus revenue, not necessarily handheld. They'll give away the handhelds. And then they'll bundle it and try to with noncompetitive ways, keep us out of the hospital. But like that's how people process. We're a one-to-one company that has individual probe and the software. And when we show people how many doctors care about this, the awareness is actually increasing now. The conversations we're having are less who are you type of thing. And that's why like we had some headwinds this year because we had some very big deals lined up, but then it was just so easy to push us aside because we're not in the budget, we're not the top priority in the hospital and -- but that's going to change.

Yes. Yes. Okay. Where are the bottlenecks in enterprise approvals? So like thinking the cybersecurity, integration, training, budget, you just touched on budget. And how do recent security accreditations shorten that cycle?

Well, so the biggest bottleneck is because like every hospital has a massive EMR implementation. They're doing something. And so to get a priority up on that list and get them to carve out the resources at times because we have to integrate into some of their systems. And we've done everything to make it as easy as possible. That's the biggest bottleneck. But there's also this, look, beautiful hypocrisy. And I don't blame them because I did this when I was working for a big company. But most of the other devices that the big companies have aren't connect -- really cloud connected. They'll connect into a DICOM and push an image, but they're not really connected. And so what they do is that they sell to their disadvantage. And they say, "Okay, well, we're this closed system and this other company has a cloud and oh my gosh, that's not going to be secure." Well, every health system has a cloud instance. Every -- there's AWS special cloud, Salesforce and Cerner and whatnot.

So they're all in the cloud. And so we've -- but the hypocrisy is if you listen to the CEOs of those companies, we'll have a cloud instance in the next 3 to 5 years. We're building AI and cloud. They're just saying, we want to be Butterfly, we want to do a Butterfly, though. S while their sales force is selling against it, they're actively developing it. And all of a sudden, it will be the best thing since sliced read and the most secure thing. So it's just a bunch of sales and marketing garbage. But it's effective when it's coming from those companies. So we've had to make sure -- so we just said, you know what, we're going all the way. And we are going to be FedRAMP certified in the next like 3 to 4 months. And that's the highest, highest security. So we'll be high trust FedRAMP, and we've gotten all these other clearances. So we know that we have very sensitive data, and we do everything possible to secure it.

Okay. Got you. You've completed development of the P5.1 chip. Am I saying that correctly, P5.1? What step change in image performance and power thermal efficiency should users expect? And how does it expand clinical use cases beyond the rent iQ3?

Great question. So it's -- we're on this journey, right? We're on this journey of more processing power, more capability, better image quality. When the first phone came -- the first camera came out, it was 1 megapixel. Apple just launched a 48-megapixel camera. So it's just like -- and we're the only company that's now doing this with a semiconductor, and we're on this like every couple of years, 2 to 3 years, we're making these generational type chases. The next version of the chip does something that no one in the industry thought can be done. What's kind of fascinating is in Europe, there's a law called RoHS, restrictions on hazardous substances. And so every electronic device, not just health care, is subject to this law. And the law, what it does is it takes toxic materials out of electronics to improve the environment.

So for the last 17 years, they've -- the ultrasound industry has had to go through a trade organization called Corsair and every 3 to 4 years, file for an exemption to this law because they have lead crystals in the ultrasound devices. And so I didn't even know this law existed and one of our partners is like, well, you guys aren't you compliant to this law? The whole industry needs an exemption, but because we don't have lead in our devices, we're compliant. So we went back and we read every one of the public applications that they made to the regulators saying that, oh, we can't do it, and the other companies that are using semiconductors are never going to get there because they can't accomplish the type of mechanical impedance. And there's a pressure that require -- that gets to a better image. It's almost like you've ever been on a boat and you just hit the gas and you start going. And then after a while, the boat kind of planes and then goes fast.

Yes. Yes.

It's kind of the same thing where when you get to a certain amount of mechanical pressure, the waveform that goes to the body, all of a sudden kind of glides through the body, and it's called harmonics. And they literally in their application said they will never be able to accomplish harmonics. And so by reading these exemption packages, we're reading what they're telling regulators can't be done. Well, we just did it.

Yes.

So we've doubled our mechanical impedance and our P5.1 chip will now get to harmonics. And so that has taken image quality, especially in cardiac. So like we proved that we can build this all-in-one ultrasound device, but we are almost sought -- looked at as a wellness device. We then just launched the most recent version, and now we're kind of equivalent to the whole industry. Next one is going to be better. And that better is the miracle, why would you buy? It's like if I have a digital camera that's better, why would you buy film.

Right.

Because you want to have to develop every picture you don't want or do you want to now move into the next generation. So that's what's happening to us with this next chip. And now we've just started development of the following generation chip that will have 20x the processing power of our current product.

Okay. And so it sounds like one of the big -- so you've gotten there on the image quality. That was something that with prior generations was not quite at parity. So we're there from an image quality perspective. It sounds like right now, it's just literally awareness.

Well, yes. Well, the biggest barrier for our growth is education.

Yes. Okay.

MRIs, what do they tell you when you go into the room, lie down and don't move, right? Then they go outside and they start an algorithm and the thing goes through an automatic image capture algorithm. Same thing with the CT. They go out of the room and they press a button. With ultrasound, it's like being in a dark room with a flashlight. You have to move it and know where the anatomy is you're looking for. And then when you find it, you have to know what the hell it is, right? And so it's the highest learning curve modality. So we've now made ultrasound devices for $3,000, $4,000. We have 100,000 of them out there, all this other stuff. What is our limit to growth is how many people can read ultrasound, how many people can capture images. And that's the biggest regulator of our adoption. What's going to uncork that?

Yes.

AI. Okay. We have now -- we created Butterfly Garden.

That's my next question. So let's go there.

We have -- there you go. We have today 27 different companies developing AI tools that will go on Butterfly. The way it works is that we've built this package that makes it easy for them to integrate into our platform. They then develop their learning models and their AI and their software on top of that hooking into our system. They go get regulatory approval, which stands on top of all of our work also. When they receive regulatory approval, they publish the app into the Apple App Store. They download the app, they contact the company, they buy a license, they have an ID and password, they plug in their Butterfly. And now they have a new AI tool.

So HeartFocus, for example, just got FDA approval. Now every user can just download an app and do a cardiac echo on a Butterfly. No integration, nothing. And there are these 27 other companies are developing all these tools to do deep vein thrombosis, pulmonary scanning, aortic scanning, abdominal scanning. And so as all these apps come out, you're not going to have to go through that kind of learning curve anymore. You're going to have these automated tools that make it easy to get out there. Look what happened to the BlackBerry, right? Secure locked down. Everyone had to do it, it was -- right? But what killed the BlackBerry?

Phones.

Apps.

Apps.

Because every new app that came out, all of a sudden, now I can get rid of the GPS off my dashboard. I can get rid of my alarm clock from the side of my bet. I don't need to actually wear a watch for seeing time other than for jewelry.

I can find a boyfriend on my phone.

You see.

Yes.

And now you have a boyfriend finding app.

Yes.

So when you were forced to use that BlackBerry, you all of a sudden went and bought a phone. And you started -- and you bought a phone because you wanted all those apps, right? That's what we're going to do to the whole ultrasound industry. We're going to have every AI tool out there on Butterfly, and it's going to -- and on top of that with our processor and our AI capabilities and now with our next chip having edge compute, so we can actually compute on the device, we're just -- we're going to obsolete this industry.

Yes. So let's talk about that a little bit because I think -- well, first of all, it sounds like what you're saying, and I don't want to overuse this, but I think you guys used to used this term democratizing ultrasound. So this is what you're talking about, it sounds like. And any nurse in the ICU can pick up one of these devices...

Check this out. So you probably remember this, but we signed a deal with Gates where we were going to.

Gates Foundation.Yes.

Gates Foundation. We trained 500 midwives in Kenya and 500 midwives in South Africa. And they're now doing something like 90,000 scans a month, and they're saving -- they're reducing birth rate complications and deaths because they can reposition the baby now because they know exactly the position. And so we taught all of those people how to do that. We now have just launched in 2 countries in Africa, I forget the countries. Malawi and I forget, but 2 countries have a tool with the Gates Foundation, which can determine -- so women, especially in third-world countries, don't know the age of the baby because they don't know the exact date of conception, and they don't have the type of rigor to get the early ultrasound.

So they find out a little bit later that they're pregnant. And now -- and many of them have to travel to go for birth, and they don't know when that birth date or when that delivery date would be. So we have an AI tool developed at the University of North Carolina that they just paint the pregnant abdomen, the pregnant uterus, and it will tell you the age of the child. It will just spit out a number. There's no like I have to interpret the image. You just have to do this up and back. And once you finish that, this is how -- what the age is of the child. Now is that a profound change to health care? Does that like -- that's like holy cow. I can totally change these health dynamics with a low-cost, portable, mobile, durable, programmable, infinitely upgradable technology that's what I'm saying.

Yes. Yes. No, that's super cool. I mean one of the things that I imagine -- well, you tell me, I imagine you get some questions around is the other big -- you're going up against some behemoths in imaging in the hospital. I mean, how do you keep them at bay?

We don't. We don't.

Okay.

We don't. They go in and they sign multimillion dollar -- I can show you all -- I have all the press releases from all our competitors.

Yes.

And what they do is they sign these research and development deals "with hospitals". They get the MR business, the CT business, the ultrasound business. And even some of them -- and they're like -- and they don't allow them to buy other people's MRs.

Other people's -- yes.

Now that's -- if we keep bumping into that, there might be a problem because it's anticompetitive. So -- and they're using that type of leverage. But no, it's -- we just the other day had a health system. We went through the whole process of getting an MSA because they want to buy it and this and that. And then all of a sudden, they get a letter, oh, we didn't realize we're not allowed to buy our product. Now to buy our product. So that's going to come to a head.

Yes, I can't imagine the hospital customer is going to let that continue.

No.

Okay. That's interesting. So let's talk a little bit more about Butterfly Garden, some of the evidence. So the JAMA published study on the Auto B-line Counter showed a 30% reduction in length of stay and greater than $750,000 in direct cost savings over 6 months. How are results like these and maybe talk about other things you might be working on to build the clinical evidence and drive the ROI message home with the hospital. How are these being incorporated into enterprise ROI models and also payer discussions?

So we talked earlier about GoScans. And so there's -- for putting the software in, there's a really strong net economic to just capture what they're already doing. But we believe, and it's proven in almost every other specialty that the early you can diagnose something, the better you can make a decision.

Absolutely.

There's not much evidence to the contrary, right?

Yes. We had an AI-focused panel yesterday just about that, yes.

Just about that. And so years ago, we started to invest in these different clinical trials or experiments to try to prove on that and try to say, hey -- and it's almost like we're trying to prove something that's so obvious, but we're doing it anyways. And in this example, they took one intensive care group who just do standard of care and then one does standard of care plus they're trained with Butterfly. So when they do their rounds, they can just quickly take the device out and just do a quick look and see what they see. And in the group, it was 200 patients, 100 in each group, all risk stratified. So it's not like there's any anomalies. In the group with the Butterfly, they were able to reduce their hospital length of stay by 30%. So those 100 patients reduced the cost to the hospital by $750,000. And that's just because they're taught to use something that every time they engage a patient, they can get data.

It's same like a pilot we just did for a skilled nursing facility where they have all of these readmissions for congestive heart failure. It's 100 patients that we did. And normally, it's 25% to 40% readmission. In the 100 where like every other week to a month that they would scan these patients when readmission at day 72. So by just doing things -- and so the essence of Butterfly going back to the first statement is every doctor, every nurse having a device, take it out, do a diagnosis, make a decision or get the data for someone remotely to make a decision. If it's a nurse who is not skilled enough to read the ultrasound, but they can capture it. So AI is going to change the world. This will be a part of every single workflow, and we have so many -- there's now -- we went and did a PubMed search, and there's like 1,800 studies a year on point-of-care ultrasound and more than half of them include Butterfly.

Butterfly.

So there's a preponderance of evidence. It's just now the hospital administrators have to make this a priority to capture those GoScans, but then also loosen their CME dollars and say, look, hey, docs, we're going to educate you because you know what, 70% of the medical schools in the United States are teaching kids on Butterfly.

Yes, I do remember that.

So that generation...

And veterinary schools.

Veterinary too.

Yes.

Especially down Texas Tech. So yes, 100%. So this next generation, like we just had in 2025, several schools graduate 4-year students with ultrasound. So those residents are now going in completely distinguished from all their peers, even their attendees. So it is going to happen. It is inevitable and that's exciting.

Yes. And just to go back to the competitive landscape here, I mean -- and maybe talk a little bit about the moat you guys have created from a technology perspective. I think you alluded to it earlier with talking about what they were telling regulators they can't do.

Well, it cost us $300 million to develop these semiconductors. We have 600 patents. And so much of what this has been -- and now we're on -- we'll be on our fifth or sixth generation -- fifth generation chip. So -- and there's so much know-how. There's so much trial and error for someone to start a program right now, they're so far behind.

Yes. Yes, I hear that. So we only have a minute left, but I do want to touch on path to profitability. And Megan, maybe you can chime in here. We'll give you a break for a second. But -- and margins on this because this is -- we are talking about pretty expensive components here.

Yes. So right now, this year, we've seen an adjusted gross margin of around 64%, which is an improvement over last year Q3 of about 4 points. That's driven by a couple of different things. One is an increase in our average selling price with us selling more iQ3s. We were able to raise the price on that. We've also seen some manufacturing efficiencies as we now have more experience manufacturing, the iQ3. So I mean, we're really excited about the progress there.

And any exposure for you guys on the tariff side of things given the components of your devices?

Sure. So our probes are assembled in Thailand, which is subject to a 19% tariff. There's not really an impact in 2025 because we had the ability to kind of manage our inventory upon the announcement to mitigate that. Going forward, had the tariffs been in place for the entirety of 2025, we probably would have seen somewhere around $1 million of impact. So there is an impact, but we're not expecting it to be...

Pretty manageable it sounds like.

Yes. Exactly.

Okay.

Well, listen, with that, we actually are at time. So thank you guys so much for sharing the Butterfly story.

And thank you for...

Fun to see the progress that you guys have made. Thank you so much.

Thank you, Danielle.

Thank you.

Good morning, and thank you for joining us. Earlier today, Butterfly released financial results for the third quarter ended September 30, 2025, and provided a business update. The release, which includes a reconciliation of management's use of non-GAAP financial measures compared to the most applicable GAAP measures is currently available on the Investors section of the company's website at ir.butterflynetworks.com. I, Megan Carlson, Interim Chief Financial Officer of Butterfly, alongside Joseph DeVivo, Butterfly's Chairman and Chief Executive Officer, will host this morning's call.

During today's call, we will be making certain forward-looking statements. These statements may include, among other things, expectations with respect to financial results, future performance, development and commercialization of products and services, potential regulatory approvals and the size and potential growth of current or future markets for our products and services and changes in the nature of our business. These forward-looking statements are based on current information, assumptions and expectations that are subject to change and involve a number of known and unknown risks, uncertainties and other factors that may cause actual results to differ materially from those contained in the forward-looking statements. These and other risks are described in our filings made with the Securities and Exchange Commission. You are cautioned not to place undue reliance on these forward-looking statements, and the company disclaims any obligation to update such statements.

As a reminder, this call is being webcast live and recorded. To access the webcast, please visit the Events section of our investor website. A replay of the event will also be available on this page following the call.

I would now like to turn the call over to Joe.

So thanks, Megan. Good morning, and thank you for joining us for our third quarter 2025 conference call. We're pleased to report the third quarter '25 results were at the higher end of revenue guidance. We continue to consume less and less cash while driving gross margins higher, excluding our noncash inventory adjustments that we recorded. Please keep in mind that our third quarter 2025 results compared to last year's third quarter where we delivered 30% revenue growth on the heels of a very successful iQ3 launch. We were able to keep growing on top of that very strong quarter a year ago.

We knew coming into 2025, we would have to anniversary a big year and new product launch. In major medical institutions across the United States, there are hundreds of doctors who own a Butterfly. To drive enterprise sales, we developed a strategy that builds on the base of individual support and introduces Butterfly more holistically through the health system. During the second quarter of this year, we encountered some headwinds in the strategy as hospitals were focused on broader macro issues. Those headwinds remained during the third quarter.

Nonetheless, our pipeline opportunities have increased. We believe we're starting to see the cloud lifting. I expect we'll return to the momentum we're used to in 2026 and may even see early signs in the current quarter. When we laid out our 5-year plan in March 2024, we introduced multiple growth pillars. One of those was our core business, going deeper into the POCUS category with higher quality imaging and software that meets the needs of hospital systems.

We also said we'd expand into entirely new markets, and we continue innovating to maintain our differentiated edge. We're seeing that our strategy is working and that our many shots on goal position us for success. We're improving healthcare economics every day and are intent on bringing this innovation directly into the heart of medical care. While the third quarter is usually our quietest, we kept busy driving progress across the business and continuing to lay the necessary foundation as we mature and scale. An important part of that foundation is ensuring that we remain a trusted partner to our customers and information security plays a big role in that.

We've always had a strong security posture, and we've now strengthened that further with ISO 27001 and other international certifications announced earlier this week. Each of these reinforces that Butterfly's cloud is safe, trusted and enterprise-ready as we expand globally. Our AI strategy is also proving to be a great accelerator for us, and it's really come to life in Q3. A major milestone came in September when the POCUS CARE trial from Rutgers Robert Wood Johnson Medical School published in JAMA, highlighted the real-world impact of Butterfly's AI lung tool, or auto beeline counter on patient care and hospital efficiency.

In this evaluation of over 200 patients, integrating Butterfly's tool into workflows, improved clinical management in 35% of the cases, reduced hospital length of stay by 30% and generated more than $750,000 in direct cost savings. This is exactly the type of clinical and economic validation that fuels our enterprise strategy and shows the power of AI to drive real change in health care. And our next-gen enterprise software, Compass AI is on track to launch before year-end. So we expect to take this type of impact further by making large-scale hospital use even easier and more effective.

We've also said time and time again that education is the biggest barrier to mass adoption of point-of-care ultrasound. Butterfly has invested throughout our journey in education from in-person training to advanced AI tools. Butterfly Garden is our ecosystem to deploy AI tools to caregivers, including those who were not classically trained in ultrasound. During the third quarter, HeartFocus from DESKi became the first FDA-cleared Butterfly partner app to launch in the Garden.

HeartFocus uses AI to guide echo probe placement, capture quality images and support faster cardiac scans for any healthcare professional. So we're excited with this launch and now it's bringing action into the vision that we've had. Echocardiograms are done about 7 million times a year in the U.S. with a wide range of procedural costs from $250 a study and sometimes well over $1,000 out of pocket. That's over $1.7 billion to payers on the lower end. Most of that happens in a facility. And if the cost isn't a barrier, access is.

Now anyone with a Butterfly can get a HeartFocus license, download the app from the Apple Store, plug in their probe and perform a limited echo, a scan that's often all you need to get key information fast right at the bedside. It's that easy, and this is why Butterfly Garden is an amplifier. It lets more users practice medicine in ways they couldn't before, improving patient access to timely care. We also have Butterfly ScanLab in our very own AI-powered app that members can use for ultrasound learning at no additional cost.

Last October, Kansas City University College of Osteopathic Medicine became the first medical school to use ScanLab in an elective course. This week, they joined us for a webinar to share their results, which 95% of students committed to independent scanning, over 7,000 scans were reviewed and more than 230 faculty hours were saved. This was made possible because they leveraged AI through ScanLab. In fact, the impact and student enthusiasm was so strong, as of this fall, KCU expanded to a one-to-one model using our probes and ScanLab across all 4 years of their curriculum.

Continuing on AI. For the last several years, Butterfly has been working with the Gates Foundation and the University of North Carolina on an AI-powered gestational age calculator. In many low- and middle-income countries, women often learn their pregnant late in their term and may not know the date of conception. Yet gestational age is critical in guiding the care and could be life-saving for those far from health facilities. This is where the AI steps in.

So using a simple set of blind sweeps across the uterus during pregnancy, the tool automatically calculates the age of the baby. No image interpretation or specialized training required. It's a powerful example of technology meeting accessibility and affordability. Since 2022, more than 1,000 Butterfly devices have been deployed in over 1,000 healthcare providers across sub-Saharan Africa, resulting in about 2 million scans to date. What's exciting is as of Q3, caregivers now in Malawi and Uganda can now use the AI calculator directly in the Butterfly app. No additional hardware needed. And with the tool currently under FDA review, we're looking ahead at bringing it to even more settings.

Each of Butterfly's AI initiatives have the opportunity to drive a paradigm shift in how care is delivered and represents exactly what Butterfly stands for, using AI and handheld ultrasound to remove barriers, empower more providers and help improve outcomes. We've built the framework and the platform, and now AI is unlocking massive leverage and bringing our growth strategy to life.

So I'll pause here and turn it over to Megan to review the financial results for the quarter. Megan?

Thank you, Joe. Revenue for the third quarter of 2025 was $21.5 million, reflecting 5% growth over the prior year period, which was primarily driven by higher average selling prices from a larger percentage of iQ3 sales internationally as well as increases in volume mainly in the U.S. Breaking things down between U.S. and international channels, total international revenue increased 4% over the prior year period to $5.4 million. The increase was driven by price given the international launch of iQ3 in the third quarter last year.

During the third quarter, U.S. revenue was $16.1 million, which was up slightly from the third quarter of the prior year. The slight gain was due to e-com sales as well as improved performance in our veterinary distribution channel. Breaking our revenue down between product and software and other services, product revenue was $14.6 million, an increase of 8% versus Q3 2024. This increase was largely driven by higher average selling prices in our international markets as well as increased volume within both e-com and vet.

Software and other services revenue was $6.9 million in the third quarter, which was flat to the prior year period. During the period, we saw increased licensing and services revenue from our partnerships, offset by lower renewals of individual subscriptions and lower revenue from extended warranties as the standard warranty of our iQ3 probe is longer than our prior models.

Software and other services mix was 32% of revenue, which was slightly lower than the third quarter of 2024. The percentage of revenue from software and services has decreased in recent quarters as our product revenue growth outpaced software revenue with the launch of the iQ3 in early 2024 as well as our geographic expansion.

Turning now to gross profit. Gross margin, including a noncash write-off of excess inventory of $17.4 million was negative 17.5% compared with 59.5% in the prior year. Adjusted gross margin, which excludes the impact of the inventory write-downs, increased to 63.9% from 60% in the prior year period. The increase in adjusted gross margin was driven by an increase in average selling prices as well as a reduction in software amortization costs.

To expand on the inventory write-off, during Q3, we recorded a noncash charge for the write-off of excess quantities of our previous generation chip that are used in the manufacturing of our iQ+ probes. We originally expected iQ+ to continue to be a larger portion of our volume. However, the strong market adoption of iQ3 has outpaced expectations, prompting us to revise our assumptions to reflect a higher proportion of volume attributable to iQ3.

In quarter 3, for example, iQ3 accounted for approximately 85% of our probe volume and iQ+ represented the remaining 15%. While iQ+ will continue to serve as the lower cost alternative and address targeted use cases, our earlier forecast assumed a greater share of demand from iQ+. We've since refined our forecast to reflect the actual product mix and market trajectory, resulting in the write-down.

Moving to adjusted EBITDA and capital resources. For the third quarter of 2025, adjusted EBITDA loss was $8.1 million compared with a loss of $8.4 million for the same period in 2024. The improvement in adjusted EBITDA was driven by the previously mentioned improvement in adjusted gross profit. These reductions and improvements led to a normalized cash burn of $3.9 million. Cash and cash equivalents, including restricted cash at the end of the quarter were $148 million, and the trailing 12-month use of cash was $31.5 million.

Before turning to guidance, I can update you on some macroeconomic factors. As of this morning, we are on the 31st day of the federal government shutdown. To date, we have not been directly or significantly impacted by this. However, we're keeping our eyes on customers that may be impacted as well as agencies such as the FDA. A shorter-term shutdown is not expected to affect our sales pipeline, but a prolonged closure could delay deal timing for the deals that rely on some degree of government funding. However, this is not currently a significant portion of our pipeline.

We're also exposed to the indirect, more systemic impacts of a prolonged shutdown such as customer cash flow timing as a result of impact to payers. As of now, we don't see this as a significant risk to our business. Additionally, at this time, the FDA has paused fee-based submissions during the closure, though they'll continue to review in-flight submissions for the time being. Again, a short-term shutdown is not expected to impact our submission time line. However, should it extend significantly, regulatory processing delays could become a factor. We will keep you updated on this matter as it progresses.

Next, we continue to see a trend of some of our customers delaying purchase decisions as they navigate macroeconomic factors. And while Q3 is typically our softest quarter of the year, contributing to this were purchase delays that impacted our U.S. hospital and enterprise channels. While timing remains uncertain, we have several large deals in our pipeline we expected to close earlier in the year that remain active.

From an opportunity perspective, in addition to unlocking the Octiv pipeline, we are working on several deals within our Octiv business and continue to negotiate an agreement with a large insurance company to reduce readmissions. As soon as we have updates on these, we'll let you know. When we evaluate our headwinds and opportunities together, we are reaffirming our full year revenue guidance in a range of $91 million to $95 million, which implies $25 million to $29 million in revenue for Q4. In order to get to the higher end of the range, we need to close on some of the larger deals that are in our pipeline.

Given the visibility we have into the remainder of the year, we are able to tighten our full year adjusted EBITDA loss guidance to a range of $32 million to $35 million or $9 million to $12 million for Q4. We have continued to maintain our disciplined approach to expense control while also investing appropriately beyond our growth areas to enhance our delivery capabilities as additional revenue opportunities crystallize.

To summarize, we delivered on the top and bottom line in Q3. And while uncertainties continue to exist around the impact of the government shutdown or the outcome of policy decisions from the administration, we have strengthened the diversification of our business and are excited about the opportunities in front of us. Butterfly is extremely well positioned to meet the needs of our customers as our technology not only enables superior flexibility and strong image quality, but has allowed us to be a much more affordable solution at scale than the current cart-based ultrasound solutions.

In addition, our semiconductor development path will continue to improve this price performance advantage with each subsequent generation. Simply put, we see Butterfly as a long-term winner in ultrasound in any macro environment.

With that, I'll turn the call back to Joe.

So thanks, Megan. As we discuss almost every quarter, we're working on a compelling growth plan that builds first on our market leadership in point-of-care ultrasound. As I mentioned before, it's coming to life through education, cloud compatibility, incredible technology and accelerated through our leadership in AI. We're also building on our POCUS business with strategic initiatives that we're getting closer to delivering in a meaningful way.

We are working every day to transition our home care pilot to a commercial agreement. And the moment we have a significant update, we'll let you know. In parallel, we're engaging with additional national risk-bearing organizations that are receptive to the model we tested earlier in the year. What we've learned is clear that selling probes isn't enough for large at-risk providers. They need scalable models that educate, manage data and competency and enable growth through new use cases and AI. I believe a meaningful part of Butterfly's future revenue will come from solutions, not just devices.

Selling hardware alone is in the past and pairing it with software, services and hands-on support is how we'll scale. I'm hopeful we'll bring the first of these opportunities across the finish line very soon, signaling the start of a potent next chapter for Butterfly. I'm also happy to share that we have officially completed the development of our P5.1 chip. As discussed during our 2024 Investor Day, Butterfly's next-gen chip was designed to integrate advanced MEMS capabilities that significantly increase the mechanical pressure associated with imaging, something critics of CMUT technology long claimed couldn't be done.

They said our digital approach would never accomplish harmonics like piezo-based handhelds. Well, they were wrong. P5.1 is now entering fab production, and we expect that in the second half of next year, it will debut in its new form factor. If Butterfly iQ3 established performance parity with other handhelds across key presets, helping fuel our sales growth over the past 2 years, P5.1 will surpass them entirely with the potential to make piezo handhelds a thing of the past for nearly all use cases.

Apple recently launched its 48-megapixel chip, a continued manifestation of Moore's Law, demonstrating they will never stop innovating. Well, neither will we. Our image quality will continue to make exponential leaps for Butterfly. The only question will be, why would anyone buy a piezo handheld anymore? These competitors are loading their devices with more LED crystals, trying to keep up with what Butterfly's all-in-one digital platform already delivers. But piezo is yesterday's technology. Soon, these devices will be in the drawer next to the film cameras. Big Piezo may be well funded, but they failed to invest and now they no longer deserve the market.

I'm equally excited to share that with P5.1's release to fab production, we transition into the beginning of our sixth-generation Apollo AI chip development. Apollo AI is designed to deliver not only lightning fast ultrasound processing, but also a local AI capability at the edge. The upcoming Apollo AI chip introduces a scalable architecture that seamlessly fuses Butterfly's proprietary ultrasound front end with the advanced digital processing, achieving a much smaller chip size and greater power efficiency in order to deliver even better image quality and capability across a broad range of clinical applications.

It's intended to support both on-device and edge AI acceleration, providing flexible compute expansion and integration with leading AI platforms. With its intelligent architecture and future-ready design, Apollo AI will serve as the foundation of Butterfly's next wave of innovation, advancing diagnostic imaging performance and enabling a new era of AI-driven medical insight.

So finally, Octiv. There are several large ultrasound on-chip partnership discussions that we are continuing having. And while I'm still not at liberty to unpack them for you, I hope you can sense my enthusiasm. Progress is happening and the time line that allows us to share news with you is largely driven by our partners. What I can share is this, if these opportunities play out the way we see them shaping up, Butterfly may, over time, transition from a POCUS company with an ultrasound chip to an ultrasound chip company with a POCUS business.

That's a step change in how our core technology will scale and how our total addressable market will multiply with POCUS becoming just one of many ways ultrasound on semiconductors can be deployed across some of the largest medical applications in the world. I am looking forward to providing more information as it becomes available and when our partners are ready.

As I close, it's worth remembering just 1.5 years ago, we set out to transform this business. Today, I can see the vision now coming to life. Soon, we'll be swimming in a big, beautiful sea of Blue Ocean.

So with that, operator, let's move to Q&A.[ id="-1" name="Operator" /> [Operator Instructions] And our first question comes from the line of Chase Knitbacker with Craig-Hallum.

Maybe just first, if there's any way you can help us kind of quantify the size of some of these deals that have pushed because of the macro and kind of what you're seeing from an activity perspective in October as far as it relates to those and the rest of your pipeline as we -- as when we look at kind of guidance, what it implies for Q4, it does imply a nice step-up. And so maybe just speak to kind of your confidence in that pipeline and the activity so far in October.

Well, yes, we've been seeing through the year deals just simply push, and we just haven't been able to get them closed. So those can be in a size of 100 to 200 probes. We have some pretty large medical school deals also. So it's kind of across the board when it comes to anything over 100 probes.

And then just as far as October has...

Yes, regarding my confidence, we really -- the beautiful thing is that we're not losing deals. This is not a competitive issue. This is not an issue where -- and I think people and especially as we griceite ourselves with hospital administrators very much for the first time, there's an inevitability to this. As I said earlier in my comments, doctors are purchasing Butterflies on their own. And when we tell administrators how many people in their institution own a Butterfly, they're shocked.

So -- and not only that, but there's such a preponderance of ghost scans, scans that are not being reimbursed because it's not a part of the ecosystem. So we definitely feel there's inevitability, and we see a lot of activity and people going into the end of the year. I think when we get into a whole new 2026 budget cycle, I think people will feel more comfortable about spending on like brand-new projects. I've seen in other companies where people are saying, okay, well, CapEx is fine. That's because it's budgeted CapEx and they're getting reorders.

To do a whole new project to carve out time from the IT department because everyone has some major EMR project. But the carve-out time in the IT department and carve-out time for a big project, that's been just the easiest thing from the delay. So we're definitely seeing that the opportunities are stacking up, which is good. And we're very hopeful -- we wouldn't guide to the quarter if we didn't feel confident with what we put out there.

Understood. And maybe just on the subs and software side of the business. It sounds like there's a little bit of some lower renewal rates. Can you just maybe discuss kind of drivers there and kind of how you think you can kind of improve some of those metrics, particularly with Compass on the horizon here?

Yes, sure. Thank you for the question. So we continue to see churn in our individual subscription as we've seen over the past couple of quarters. And also year-to-date, we've seen an uptick in our enterprise subscription. In terms of timing of subscription as a percentage of revenue, a lot of it is just that timing because the software and the hardware have different revenue recognition patterns. But as you mentioned, we are very excited about the rollout of Compass AI later in Q4.

[ id="-1" name="Operator" /> And the next question comes from Josh Jennings with TD Cowen.

I wanted to just follow up on the sales funnel or pipeline and just ask about -- maybe remind us of the sales cycle and timing and just how the pipeline is shaping up for 2026 in terms of an outlook for growth. It sounds like with some of the delays -- with some of the concerns or just tightening of capital spending for new projects that the sales pipeline may be more full heading into '26 than what you're experiencing heading into 2025, but maybe just help us think through that as well.

Yes, Josh. Well, clearly, the time to close has extended. So we have -- we definitely have a lot of deals that have aged beyond what we would normally have an age. So if anything, we're stacking up more and more deals, but we haven't been closed fast enough. So again, I think that's going to -- that will lift in '26. I think there was a bit of a shock to the system in '25 and then people have been mitigating and managing it.

But whenever we get to the end of the year, there are always dollars that people have to spend, and we'll be fighting for those dollars.

Just on the -- just the Robert Wood Johnson and the cost savings and cost effectiveness, I think that cost effectiveness data is accruing nicely for Butterfly iQ platform. But I mean, how powerful can that be in terms of driving stronger interest and closing of deals as we get into 2026 as your team is able to market that Robert Wood Johnson study published in JAMA and others?

It's really important for a bunch of different reasons. The first is that it establishes that pulmonary congestion is a very key endpoint for congestive heart failure. So from a pure clinical standpoint and which -- whether we go in home, we go in outpatient or we go in inpatient, using ultrasound and using it as a marker for congestive heart failure progression or lack of progression is really important. I'm dealing with a chest cold here.

So also, when we have our enterprise conversations, the primary place that we start from an economic side is go scans. 35% of the scans in a hospital in point-of-care ultrasound, make it all the way through to reimbursement. When they put Compass in place, that goes to 70% to 80%. So we have that great economic benefit. But that's also something that they capture that benefit in the first year of putting the software in and then going forward, they can say they're anniversarying that benefit, but where is the next benefit.

So the next big economic piece of evidence that we have identifies the fact that by using point of care in the intensive care unit, by every day a doctor just pulling out a probe very quickly, doing a pulmonary scan and using a pulmonary scan with our AI shows significant economic benefits in the ability of managing patients in the hospital. So it's not just about early diagnosis. It's about taking care of the patient's wellness at the moment that the doctor sees that, giving them the tools to see something instead of having to order a scan.

So when we sit with administrators now, we have a second big indisputable set of evidence around the economic value of doing this. And when we sit with administrators, it's not necessarily people saying, should we do this? I think actually, with this next piece of economic evidence, people believe they need to do, it's when. When am I going to take on this project? When can I stack it in with IT? When should we move forward? And so it's just such a wonderful piece of evidence that helps our economic argument when we think about our enterprise sale.

Excellent. If I could sneak in one more. I mean it sounds -- I know there's no update on the home program today, but you did complete a pilot program with a major partner and payer. Just to check the box. It's not a matter of if, but when it sounds like, and you'll just provide an update when that partnership kind of finalizes and no diminished kind of optimism or enthusiasm about that channel and that opportunity is my understanding, but I just wanted to clarify that.

Well, not certainly about the channel, but until you have ink on something, it's not real. So I would have liked to have had it done by now. It will be done when it's done. I just don't control it.

[ id="-1" name="Operator" /> [Operator Instructions] Our next question comes from Andrew Brackmann with William Blair.

So you seems very confident on the P5.1 chip and new form factor launching next year. Can you maybe just sort of talk to us about what's needed to get to the point of launch sort of between now and then what you're going to be working on to sort of derisk that launch? And then when you get there, anything you can share on just sort of the product build ahead of that launch or anything on pricing that you might expect for P5.1?

Well, thank you. So P5.1 will certainly be a highly specialized product. We haven't identified pricing yet. We'll get to that closer. As you saw with iQ3, we've been able to establish a new price watermark. We've been able to improve our overall corporate gross margins. We've been able to maintain not only a higher blended ASP, but at that higher ASP, iQ3 has outsold iQ+ to a pretty significant magnitude more than we expected.

So we -- it's not implausible to think those trends would continue with the next version. Putting a chip in a probe, tuning the chip to our software and AI, getting that probe completely -- these -- every year, we launch a new technology or develop it, it's a great coalescing effort for our company because there's not a team in the business that's not touching that new product. So iQ3 was a very successful launch. It's a very successful effort on getting the technology to market. And I don't think there's any different.

There's -- the beautiful part of when we're bringing in new processors is we're not having to reinvent every wheel. We bring a new processor in. We tune all the components and all the software to the edge capability of that processor and then we roll it out. So I think the risk to execution on launching the new product is pretty low. I think there's always risk. The call it research for that purpose. But I think right now, we look really good.

That's perfect. And then just on the info security piece, you had the press release out earlier this week and then obviously touched on this call. Maybe just in practical terms, how does this help you win business? And as you sort of think about sort of what you've achieved here, do you think that this is sort of table stakes for the entire market? Or is it a sort of meaningful differentiator for Butterfly?

It's an interesting dynamic because aside from one other independent competitor, and I could be wrong maybe by one. But in general, the industry is on-prem. The industry has software that they'll connect to a WiFi and they'll push an image into a DICOM, but they're not cloud connected. And so when you're not cloud connected and you're on-prem, you don't really have to worry about security because people don't -- can't get into your device. But our competitors have clearly communicated that they really want to follow in our footsteps and finally catch up to where we've been for a while.

And so the industry is going to cloud connectivity on imaging devices. And so historically, our competitors would use it as a selling disadvantage for Butterfly. They would say, oh, you have to connect it to the cloud. So you have to now have all of this security concerns, which is kind of garbage because every single hospital has an AWS cloud interface. And so we have -- by being the first company that's mass commercial in the cloud, we've had to kind of set the bar on security and make sure that we deal with data compartmentalization, we deal with the HIPAA requirements proper. We deal with all the different multiple layers to ensure that the data is secure.

Because remember, we have 25 million images in our cloud growing 30,000 a day, keeping that data secure, making it available to our customers, porting it into their systems for the EMR, for DICOM, for reimbursement is essential. And so we expect in 2026 that we'll have our HITRUST certification and FedRAMP, which is the highest certification of any security basically worldwide. It's the most complex.

And so to have a cloud environment, have that level of security is insane. So it has always been our competitive advantage to have cloud. It's always been our competitive advantage to be there. All others will say bad things about us and say until they finally get there and then they'll say how great it is, which I think is absolutely hilarious. But we absolutely believe that cloud is the way to be able to manage large fleets to be able to manage data, push and pull large complex models, create overall connectivity. And we have literally the best cloud security posture in the industry.

[ id="-1" name="Operator" /> [Operator Instructions] And our next question comes from Ben Haynor with Lake Street Capital Markets.

Joe, on the commentary on Apollo AI, I was just curious, have there been kind of upgrades to the original design as far as the kind of edge AI aspects to it beyond what was kind of discussed at the Investor Day last year?

Yes.

Any more color there?

Yes. So we have pushed the boundary of Apollo to include the ability and to make sure that we are capable to do local AI on the chip. That is the next biggest trend in AI instead of having to calculate and process AI in the cloud and then push the devices to be able to do it actually on location is the next biggest processing feat and it creates a significant amount of local benefits in speed and capability.

Instead of having to take something, push in the cloud process it and wait for it to come back by having that processing power to do it dynamically in your hand is a major step-up. So yes, Apollo has 20x the processing power of P5.1 and our legacy. And so we're making sure that, that processing power is put to the right use.

And I guess, maybe dovetails a little bit with your last comment regarding the cloud in terms of kind of pushing and pulling models that you would be pushing and pulling the correct model or maybe not -- the relevant model to the device if you're doing XYZ scan.

Well, yes. I mean, today, one of the benefits of our cloud architecture is we do -- like it does work like an ecosystem. You have the device, you have the app and then there's this dynamic communication with the cloud. And we kind of purpose data processing where the assets are available. But as AI models get more and more sophisticated, it's certainly possible that, that becomes more of a lag and you want to have greater local control.

So for the individual device to have that kind of processing power creates a much faster way of delivering that feedback. And that is on -- literally on the edge of where AI and compute are going today.

Yes. So specialized model to push the device eventually. Got it. Go ahead.

Yes. So more and more AI capabilities be processed on the device.

Got it. And then just curious on any updates to -- with regard to IQ Station or the RoHS situation with the European Commission?

IQ Station is actively in development. And when we get closer to a date, we'll put it out there. a big part of our future strategy. And regarding RoHS standards, it's on August 1, they closed the book on any type of submissions from ourselves or our competitors. And it's in the midst of a third-party review. That third party is reviewing data and then asking questions as they see necessary.

And then hopefully, sometime next summer, they'll render their decision or their opinion to the government to the governing body, and then we will -- they'll make a decision. But it's kind of their -- all the data now is sitting there. They're crunching the data, and we're in that quiet period unless they ask questions, and we're waiting for an answer.

[ id="-1" name="Operator" /> And that was our final question. So I will hand back over to Joe DeVivo for any final comments.

So everyone, thank you very much for the support. Sorry about my seasonal cough I keep getting. We're thoroughly excited about the progress that we're making, and we're thoroughly excited about what's to come. We definitely think we had a slower start to the year, but everything that we're building is very, very strong. And I just can't wait to be able to unpack a lot of other things for you as they mature. And also, I'm just very appreciative for Megan Carlson, our Interim CFO, for doing just such an expert job in this interim. So thank you, Megan, and thank you, everyone, for your support.

[ id="-1" name="Operator" /> Thank you, everyone. This concludes today's call. You may now disconnect. Have a great rest of your day.