Cognizant Stock

I'm Arvind Ramnani, AI, digital platforms and IT services analyst at Truist Securities. Thanks for everyone sitting in. Along with me, I have CEO of Cognizant, Ravi Kumar. Ravi, thanks for taking the time.

Sure. Thank you.

So just a quick introduction, and then we'll jump into questions. Ravi became the CEO of Cognizant at a really interesting juncture of the company and a crossroads of what was happening with IT services. Since taking over, he has made AI central to how Cognizant works with its clients, supported by over $1 billion investments in GenAI. And I think it's probably worthwhile mentioning Ravi has been named to TIME AI 100 list for the past 2 years. I think the reason this is particularly important is because that list had a 75% turnover. So you just have like 25 folks that have made it 2 years in a row and I think like something to be proud of for yourself as well as Cognizant as well.

Thank you. Thank you.

So look, we have had a number of conversations over the years. I have lots of questions here, not enough time. So we'll kind of just jump right into the biggest debate for IT services. So for decades, right, this has been a fairly, I would say, well-run industry. The industry had a very predictable path to growth. You add people, more billable hours. But now with AI, kind of the opportunity basically moves very differently. What is already changing at Cognizant's delivery model because of AI?

So Arvind, thank you for hosting me. Look, the first principles of the tech services industry have to be reforged because this is a very different tech disruption. And let me tell you what that reforging means and what we have done at Cognizant. The first and foremost, this is a business which is no longer going to be about human capital. It's going to be about human and digital labor coming together to deliver a tech-led transformation. Tech-led transformation for companies like ours was about building systems. In the '90s, we built custom bespoke systems, then we built -- then we became a system integrator for classical software. Then we actually became a services player for technology and software, which was on the cloud and the plumbing was on the tap and the building was done by us.

Fast forward now, I think this is a technology which is action-oriented. It mimics human labor. A lot of what we're seeing today is productivity-led. It's still -- we've not got to a point where we do new things in new ways. We're doing old things in new ways. So the first and foremost, the first reforging of the first principles is we make the execution of this technology, not just into systems, but also business operations of companies. This market was $1 trillion before. Building systems at a time when enterprises were globalizing. This market now is $6 trillion because you actually have a total addressable spend around business operations of companies and you do it embedded with digital labor. So that's the first reforging of the first principles of this industry.

The second reforging of the first principles of this industry is to go more to an outcome-based business. I mean, enterprises across the world no longer want to buy software and no longer want to buy services separately. They want to bundle this and own -- they don't want to own the agency of outcomes. They want to transition this to a provider who can do that. The third is it's going to be a platforms plus services business, which I believe is something we have been doing for years now. In [ 2017 ], we bought this company called TriZetto, and now we manage 200 million lives of insured health care in the United States.

So platforms, human plus digital labor, outcomes driven are the reforging of the first principles. If we reforge this, we see a bigger market than ever before. It's no longer a $1 trillion market, it's a $5 trillion market. If you don't reforge it, you just end up being doing software engineering, which then means you're doing it in the old way and you're going to deflate that model. But if you expand yourself into the business operations of companies, it's a 6x opportunity.

Yes. That's a great picture, right? Like if we just kind of get this down to specific customers, you obviously have a range of customers, some are more, I would say, kind of advanced in how they're using AI and how you're helping them. Can you talk about one of those customers? And if you can name -- specific name, but maybe the industry, I think that will help.

Absolutely. So in 2024 and 2025, I would say it was the first chapter of AI, much more broad-based. It wasn't as nuanced. We dug a hole and a basement with the same machine. Most of it, the world is in a flattish growth trajectory. A lot of it was productivity led, and we kind of swapped wallet share to transfer the productivity to clients. And if you could do that efficiently, you could actually win more wallet share with clients. It doesn't mean new spend cycles, existing spend cycles done more efficiently. That's what happened in the first wave. I would say later part of last year, we saw a second swim lane triggering off. That second swim lane was doing old things in new ways. Say, for example, for a large telecom client in Asia Pacific, we are migrating their mainframe into public cloud.

And the idea of doing that is you could take the MIPS down $1 of -- a line of code in COBOL used to cost $10 to refactor. Now it costs $1.50 to refactor. And you transition that to public cloud and the money actually gets refactored from a different spend pool to services as well as to cloud providers. It's actually old things being done in new ways. Recently, we set up a campaign for a frontier finance organization. In this swim lane, it's not new discretionary spend cycles. It's old spend cycles, reorganized to give us momentum and runway. Frontier finance is the ability to run your finance function using a frontier first model.

So every Fortune 500 CFO spends 70% of their effort and time on controls and 20% to 30% on growth. If you can flip it, you could actually orient finance towards growth, you could orient finance towards more real time and you could power a finance function, which is driven on growth imperatives rather than controls. Again, you take an old thing, it's done in a different way, you do it in a new way. One of my favorite campaigns in recent times is pharmacovigilance or drug safety. I have 4 deals now where we have drug safety, which is called pharmacovigilance. We compete with clinical research organizations to win that business. We used to do a small sliver of work, which is mostly contact center-driven. And now we could own the end-to-end cycle and deliver better insights, faster life cycle, faster shrinking the speed at which you could do drug safety at a lower cost. This is a spend cycle our clients were already having. They were not just having it with us. They were doing it with clinical research organizations.

So how do I make a function frontier-led and create a new spend cycle for us versus the clients are still spending in the same vein. So this is a swim lane, which has got activated, and it is creating significant traction. So you layer it with the consolidation, then you layer it on top of it with old things in new ways and then doing new things in new ways, which is primarily -- which kind of cater to growth imperatives of organizations. For example, for a wealth management firm in the U.S., we are building agentic work around their independent wealth advisers and actually deliver more capability to those wealth advisers who are actually on a subscription model with this company. So it is a new discretionary spend cycle.

Now if I go industry by industry, Arvind, financial services is activated on all 3 swim lanes. And that's why for 3 quarters in a row, we have sequential -- we have year-on-year growth, which is at double digit. Now if all the industries activate this, you would actually see the industry back at double-digit growth. I mean if you -- if somebody used to ask me a question, you seem to be very optimistic, why isn't the industry growing at double digit. The only reason why it's not growing at double digit is the first swim lane is deflationary. You actually consolidate when you consolidate, you are passing on productivity to clients, which means it is going to be deflation to revenues. The second swim lane is new spend cycle coming to tech services. The third swim lane is discretionary led, which is clients actually spending more because they're going to -- they're planning to do new things and they're planning to do things which are related to growth using AI. If all 3 are activated, you can layer this to double-digit growth as an industry.

Yes. No, that makes sense. I mean my own research actually suggests something a little bit different, which is like some in the industry are not going to make it to the other side, right? So when do you -- I must say like someone like a Cognizant that's kind of, leaned in, like, fairly hard, right? When you look at your commercial licenses with the frontier models, Anthropic, like all your 350,000 employees have it. That's not the case across the industry. So could you make the case that when we -- when you see the final flip, like everyone doesn't win, right? Everyone doesn't win to get...

So this is moving so fast if anybody is telling you they've figured this out, I mean, they're not telling you -- they're not giving you the full picture. Here is my end state, which I can think of. The baseline is you need deployment capacity on the other end to embed frontier intelligence into everything you do in a company. Cognizant has 18,000 cloud-certified architects, but that's not my North Star. We are the largest on the planet. 18,000 cloud-certified architects, which is the highest level of certification in Anthropic. I'm just giving one example. In OpenAI, we have like 6,000 Codex badges. That's the largest again in the world. In Gemini, we have 5,000. All of that is not my North Star. My North Star is bend the cost curve at scale and flip this model on its head and get the 300,000-plus employees in the company frontier-led. That's my North Star.

So that's the basic -- that's the baseline on which I'm running. That capacity is going to help me deliver frontier intelligence into enterprises. Now we have -- we are absolutely convinced the capability is out here, the production value is here. The gap between the capability and production value is the bridge companies like ours to do this. I mean it's a tale of 2 cities. On one side, we want to pace the frontier, which is a talk in the last 1 week because the frontier is very dangerous. On the other side, we're dealing with the frontier can't do basic things because it's not grounded in the context of a company.

So I think you need that capability. And as long as you bend the cost curve at scale, companies like ours will stay relevant for the future. I mean, go back to any disruption, including the digital wave, you had boutique digital capacity delivered at high cost. I used to work for 2 such companies. They don't exist today. I mean I used to work for a company called Cambridge Technology Partners, which was an absolute pioneer in digital technologies. It doesn't exist because it couldn't bend the cost curve at scale, which is what clients need, which is what enterprises need.

Now the second important thing, as this technology evolves and as we learn through this process, between the intelligence and the enterprise, there are layers of value we can capture. Every time there is a disruption, that opportunity presents to us. We don't capture that layers. We always delegated ourselves to be a system integrator. Now we could capture those layers along with the deployment capacity at the end. At the end, the deployment capacity is like a services company. But the layers of value are because this is a bespoke opportunity, we could build the trust layers, we could build the context layers and we could also build the harness around it.

And I was telling Arvind offline, the harnesses built today, which are available today in the market like Cursor and Cognition and all, they are for software engineering. The harnesses you could really build could be for business operations because that's the universe. And what can you do that? You could capture the context so that the next time you come in, the context is distilled enough, and therefore, you're efficient in using your frontier. You could do the model routing so that you could straddle between the most expensive frontier to the cheapest open-weight model. You could coach the human with frontier intelligence live as they're doing the tasks so that they are not applying their mind on what actually transfers to the frontier and what remains with them.

And you could create a compounding factor at the end by leveraging the network of clients who deliver work, specifically for things, your clients don't mind sharing their alpha because it's not core to their business. Say, if you're doing accounts payable or you're doing procure to pay, it's not core to people's companies' businesses. It's not core to company's alpha. So they would be willing to pass it on to get the network effect. If you're doing health care operations or you're doing underwriting, that is core to your business. You may not pass the alpha.

So our endeavor from the intelligence to the enterprise, we have deployment capacity here, and we have layers of value. Some we can capture, some we can try, some we will not be successful. There will be boutique AI ecosystem players who will come into the mix. There will be hyperscalers who will also absorb these layers and some of the frontier models -- some of the frontier model companies will also build commercial value proposition because they want to capture the value because they don't have an ROI for the kind of investments they're making.

So our endeavor is to not just hold the capacity at scale, bending the cost curve, but also capture some of those layers of value. And today, we are building a trust platform. We have an agentic harness for business operations and for software engineering. I'm more optimistic about business operations. We are also trying to build one for physical AI, which is going to be the next wave. And we have built extraordinary muscle on context engineering, which is probably the new class of writing code in businesses. If writing code was a craft, the new craft is embedding the context into intelligence so that intelligence is applied better to businesses. So that's the journey we are going through. And therefore, I believe this business is going to be a platform plus services business.

Now how far have clients gone on the outcome pricing? I mean, that's an evolving science. Since 2023, since I have come on board, our time and material business, which used to be roughly 52% and our fixed price plus outcome-based business used to be the balance. We have flipped that number completely. And right now, our time and material business is only 42%. And the rest of it, 58% is fixed and outcome-based. And outcome-based has a spectrum all the way from transaction-based pricing to outcomes.

So it's an interesting time because the first principles are getting reforged. It's also an interesting time because it's not just -- labor is not the only input to your business. Labor is one of the major inputs to your business and intelligence, Frontier intelligence is embedded into it. How you can build the economics around it, how you build a craft around it, I think, is very important because our clients are no longer buying -- or rather, they have an intent to not buy intelligence directly. They want to buy it through us, which means we have to be efficient enough in embedding that intelligence and also routing it to the right economic models embedded into our work and take the risk and manage the return through that process.

Yes. That's a fairly comprehensive answer. Just a couple of threads on that, right? One is you said you kind of -- your fixed price is kind of moved to 52%, like, 58%...

It's an outcome-based.

Yes. And within that you said that contains many different segments. But the segment that's basically kind of tied to AI, whether it's the 18,000 cloud kind of 18,000 code engineers or the 6,000 Codex engineers, the folks that are basically trained and who are extensively using AI tools, is that bill rate like a step function different or just like a slight premium?

So, I mean, it's interesting. On one side, we have outcome-based pricing. And I put this chart of cost observability on the Y-axis and outcome observability on the X-axis. And you always think the upper right is the best one. The upper right is not the best one here. Clients actually look for the best suitable candidates for outcome-based pricing are high outcome observability and low cost observability. Interestingly, if the observability on cost is very high, clients don't want to do outcomes because they can see the cost.

So you could do an outcome-based pricing, but you're not going to get a nonlinear return. If the cost is not observable, clients want to leverage that. So it's interestingly, outcomes have to be on this side, observability on cost has to be lower for clients to actually say, I want to do outcome-based pricing. If the observability is high on both sides, you will do a fixed price deal. And fixed price deals are not outcome-based. Sometimes fixed price deals can actually be more -- can be less margin accretive because the observability of cost is so high that clients will not pay you more than what you should and you still take the risk.

So outcome-based pricing starts from a spectrum of managed services to transaction-based pricing to owning the outcomes. Owning the outcomes means, say, for example, in our TriZetto business, we went from selling perpetual license to subscription-based license to transaction-based pricing to now our clients are saying, why should I pay you for transaction if my transaction is auto adjudicated with AI. So I should be paying you on a number of lives you manage for us. So number of lives managed to us is actually on the other side of the spectrum, which then means I'm actually managing a number of lives versus number of transactions underneath it.

So it's a spectrum all the way and getting there is going to be a long haul, but we have spoken about outcome-based pricing for like 25 years. This is the first time we've got a chance to own it. And because we're doing operational work more than software engineering, we could own it. I mean I could go to a company and say, I will own your Know Your Customers' process or I could own your accounts payable process and give me a cut of it. A good example, we work with one of the largest food distributors where we manage their agentic process of credits. I mean, food distribution has a low margin. And the credit cycles, I mean, when you return your goods back because you don't like them or they have not come to your expectation, the money flows back to you, the goods go back and the money flows back. It's a long cycle. We could shrink it and increase your working capital efficiency. And if we do so, we can take a cut out of it. Clients have not gotten that far. They still engage with us on a fixed price model. Over a period of time, that would happen.

Now coming to time and material, interestingly, that's becoming outcome-based as well. I have an AI-infused rate card, which I introduced last year, from A0 to A4, A0 being all human effort, A1 being human effort validated by machines, A2 being machine effort validated by humans, A3 being all autonomous effort. And we presented it to a few clients. Clients loved it. They gave us a premium on the rates as you go from A0 to A1 on the human labor. Then our clients came back and they said, "Wait a minute, I thought I got a great deal on this, but the inference and the pretraining costs are on the tap. And that is not controlled. I'm paying directly. If I put all of it together, I don't think I'm getting a good deal." So now clients are saying, take the inference cost, take the pretraining costs, embed it into your billing rates and tell me what your new billing rate is, which means I have to give a billing rate for human and digital labor. If I have to do that, then I have to have traceability of the tasks, which these people are going to do so that I could measure the pretraining and inference costs.

So it's a craft we have to build so that, that then becomes the opportunity to arbitrage with clients on expertise. It's not yet happened, but clients are starting to think about -- I mean, the A0 to A4 rate cards have happened in many of my clients. So you don't give the rate card based on experience, you give rate card based on whether it is fully autonomous or fully manual. But embedding digital labor is one more level of complexity, which means the tap on Anthropic and the tap on OpenAI is going to be through us. By the way, we have a prebuy arrangement with all the 3 frontier model companies. There are deals where we have told our clients we will deliver this outcome. We went public on one -- particular one called Travelport, where the client is actually delivering operations and software engineering embedded with the frontier model. They are not paying the frontier model company. We are paying the frontier model company. So we have started to do that kind of work, but it is a craft we have to master on so that our clients give us that work along with the Frontier intelligence because we can do it better than them.

Yes. So if we can just kind of continue on the thoughts, right, where basically clients can very well go and buy these directly from an OpenAI and Anthropic. They know how to reach them. Like what's the value proposition? Or what is your value proposition as Cognizant to say like, well, if you buy from us, then like...

Yes. So essentially, you're not buying capacity from us. You're buying an outcome, you're buying to deliver a task. We are figuring out how much of digital labor, how much of human labor. And if the digital labor should come from an open weight model, come from a cheap frontier, come from an expensive frontier, which means they are not actually managing that risk and they're managing the outcome. So don't come to us because we have wholesale capacity on Anthropic. They can go directly to them. Come to us because we know which frontier to use to drill the hole and which frontier to use to drill the basement. And we know where to use a frontier and where not to use a frontier and allow us to manage that and focus on the task accomplished. That's the thesis. And that's why we have prebought this capacity so that we can deliver to that outcome and deliver to those tasks.

Now software engineering is relatively easy because it's very mature. Business operations is not easy because it is evolving. So I would like to do it because I want to -- I want to master the craft. We have to pick and choose what business operations make sense for us. Health care operations make sense for us. I mean, I'll give you a good way to say which flows are more amenable for agentic. I call them bounded workflows. Bounded workflows have 4 principles; structured inputs, measurable outcomes, high transactions, short feedback loops. If it ticks all these 4 boxes, I think I can embed intelligence, measure it well, deliver to an outcome and take the end-to-end responsibility for it. So that is what we are looking at.

So we are looking at health care operations, billing systems, high transactions, structured inputs, measurable outcomes, short feedback loops, mortgage operations. These are solid opportunities. On the horizontal side, legal operations, financial operations or the CFOs function or the frontier finance function, these are solid opportunities. So that's how we are kind of at least learning through this process.

Perfect. Got it. Terrific. Kind of shifting gears a little bit. You have a fairly, I would say, 360 or comprehensive relationships with like the Anthropic, OpenAI, right? On one hand, you have kind of the largest or the second largest footprint of Anthropic, 350,000 licenses. I think maybe Deloitte signed one, but I don't think they fully ramped. But as of now, you're probably the largest footprint. On the other hand, you have these folks who are like coded and basically kind of experts, maybe this is a smaller set, but fairly advanced in what they're doing.

At the same time, like you have some customers who are also wanting to buy and so you're kind of generating a lot of revenue through them. And you have this going on where you're essentially having more of a customer type of relationship. And then at the same time, Anthropic is out there basically building their own deployment layer like -- and clearly kind of the classic like frenemy sort of situation, like where do you think this relationship goes in 5 years?

Look, I was at one of these conferences where somebody from a frontier model company asked me this question saying, are we going to compete with you at deployment capacity? I said absolutely no. On the contrary, you're making the world believe that deployment capacity is needed. On the contrary, you're making the world believe that this deployment capacity is needed in so much value, the capability is here and the production value is here. As this progresses, bending the cost curve at scale on deployment capacity will be tested, and we will be on the front of that. So having the boutique capacity, frontier capacity from these companies, including private equity companies, which have floated these joint ventures, I think is a great thing because it just fuels the need for services.

Remember, everybody put money into the intelligence and then they figured out there was very little economics to put value in the layers before it hits the organization. So they couldn't ground the technology on the -- they couldn't ground the technology in the hustle and the heterogeneity of an enterprise, so they couldn't get value. So now that the intelligence is more ROI driven, that value is drifting into the trust layer, it is drifting into the context layer, it is drifting into reinvention of the process. I mean, reinvention of the process is real. If you are not reinventing the process, you are not reinventing the model -- operating model, there is no way you're going to get the ROI from this.

So the more this happens, the more I believe the need for frontier capacity at scale -- at a lower cost point is going to come into picture. And then companies like ours will have more mainstream role to play in this new transformation. So this is already happening now. I mean everybody is backtracking on this first chapter where they did the science experiments. They invested heavily on generic applicability of AI to now specificity, more ROI-driven, more economics and all of that is now coming into picture. So I actually feel it's a great thing. I mean it just gives endorsement to an important constituent in that layer stack, which is deployment capacity.

Terrific. I do want to ask one sort of like tricky question that everyone's been talking about over the weekend where essentially AI is going to -- the chance that AI ends up killing us all. I don't know, people are saying like a 10% chance. I think last year, what we heard from Anthropic and OpenAI is like, I'm kind of afraid that in 6, 8 months, there's going to be a massive like job destruction. Now it's going to be like, well, jobs, they backtrack from that. Now it's like, okay, maybe like not jobs like AI will just kill humans itself. Kind of what's your sort of perspective on that?

My own take on this, I mean, you don't get into the ring and then say, I can't take this to the finish line because it's going to be ruthless. The reality is pacing the frontier is what this whole thing triggered off with a bunch of employees writing and then Dario putting this [indiscernible]. And depending on who you talk to, they would tell you to pace the frontier or not pace the frontier. The reality is the duty of care is your call. You own the frontier. So you should decide whether you want to pace it or not pace it. If you don't pace it, it could be dangerous. If it is dangerous, the economic consequences of the liability you carry is going to be yours. If you pace it, somebody else is going to advance it. And the world is not in a place where everybody is going to come and there's going to be a commonality of purpose and there's going to be an agreement of what the pacing should be and what you're going to slow down on.

I mean, if I have a Genie in my pocket, I'm going to say nobody else should have one. So we don't -- I don't think we have the option to take a call on pacing the frontier or not. But if you make the duty of care as a responsible and a trustworthy duty of care and then your clients will come and buy from you because that's your duty of care, that's your calling card, then everybody else will actually put that calling card and the industry will already baseline itself with its guardrails and pace it in the way that they can control the guardrails. This is a general purpose technology. It will improve over time, and it will spawn downstream innovation.

So it's up to you to improve and control rather than saying I want to control it. Who else in the world is going to agree to control when you have advanced and the others are yet to. And it's in the hands of a few players versus an ecosystem which has access to it. And therefore, it's a level playing field for everybody else to decide that pacing the frontier is the right thing. Of course, there should be regulatory frameworks attached to it, but I don't think the players decide whether -- the players decide everybody should pace. If you want to pace it, you should pace it yourself, and you should be accountable for -- I mean, do the balance between capability and responsibility and predictability and reliability. You have to set the balance for yourself. You have to set off the capability to decide.

Now the ones who don't have the capability are going to run the extra mile to get to that capability. And the ones who have the capability will be worried whether they're going to be accountable to the liabilities of the consequences of what the danger of this technology could be. So I think this is a debate you can't win. You could argue the debate on both sides. There has to be regulation for sure. But it's the same set of companies who actually painted the doomsday situation. I mean in the United States, the anxiety on this technology is higher than the excitement. In China, the excitement is higher than the anxiety.

So we created this situation. So we have to come out of it, anchoring this to more jobs, anchoring this to what AI can productively do like cancer care and cancer cure and unlocking material sciences and stuff like that and curing disease. I mean those are extraordinary North Stars to deal with. But it's our opportunity to figure out how do you advance and control so that you get the best out of this technology.

Perfect. Just a little bit over time, but just one final question. If you look ahead over the next year, what's the one sort of metric we should -- investors should be most excited to be tracking?

Any metric you take, first of all, the most important metric is growth. I mean, if the industry was at double-digit growth, nobody would be asking me -- tell me the proxy for growth. The industry is not at double-digit growth. Financial Services, at least for Cognizant, is at double-digit growth for 3 quarters in a row. How do you -- I mean, AI revenues, AI projects, frontier capacity, all these are, I would say, not baselined metrics because everybody can paint everything as AI. Everybody can say everybody is frontier capacity. So it's hard for investors to believe in those metrics.

Now if I were to go back to the principles, which is human plus digital labor, platform play, outcome-based play, the 2 metrics which will be tangible and measurable and baseline for everybody is what -- how much percentage growth you have had on revenue per person and margin per person. If that is a -- I think that's a reasonable metric to measure. And hopefully, the industry goes back to double-digit growth at some point of time. And you're kind of anchoring this as the journey to get to double-digit growth.

Terrific. I have another 10 more questions, but a little bit over time, but thank you so much.

Thank you for the opportunity, and thank you for a very thoughtful conversation.

Thanks everyone for listening.

Welcome. I'm Bryan Keane. I cover IT services here at Citi.

And we're excited to have Cognizant here for a fireside chat. We've got Jatin Dalal, who's the CFO. We're going to run through a list of questions, and if you have any in the audience, feel free to raise your hand and we'll run a mic or I'll repeat the question for you.

But first, Jatin, thanks for coming and thanks for being here.

No, thank you. Thank you for hosting us. I appreciate the opportunity.

Yes. So I wanted to start kind of high level thinking about the IT services industry. It's growing revenue well below its historical norms. How much do you think can be explained by the geopolitical turmoil versus the secular industry pressures from AI?

I think it's an interesting question to start. If you see, there is -- the industry has seen now more than a few years or 2 or 3 years of slow growth. In our assessment, it is partially the secular pressure, but it's also the lack of discretionary spend. Because for example, in BFSI, Cognizant grew double-digit in quarter 2. And that was notwithstanding the secular pressure because of AI-led productivity and everything else that is going sort of around the world.

So I think it is some amount of secular pressure, but it is also largely the lack of discretionary spend in the rest of the sector which is sort of leading to a low growth phase for the industry as we see.

So how would you characterize the demand environment and the discretionary spend environment, demand environment in general, from maybe last year to the beginning of this year to over the last month or 2?

Absolutely. So the BFSI sector continues to be a robust performer, robust enabler or robust driver for the growth for the sector and certainly for Cognizant. We have a slightly differing sort of situation in rest of the sectors, and let me go one by one.

On CMT, which is Communications, Media and Technology, we see excellent demand from technology customers. They understand this very well and they're investing in their future. So we see excellent, sort of almost as good a discretionary spend environment on technology side. But you don't see an overall growth there because communications sector and a couple of customers, specifically one that has impacted Cognizant, has remained -- has sort of made the overall aggregate number more flattish.

Products and Resources are impacted by the geopolitical situation where they are looking through their supply chain. So there is a little bit overemphasis on today's operation versus investing on new technology or new spend. And in some form, Health is going through its own sort of policy-related predicaments or opportunities and challenges. So we see these 3 sectors in a differing space of -- or time as discretionary spend, so far as discretionary spend is concerned.

Got it. And any change you're seeing just recently, the most recent like month or 2? And going forward, does that expect -- do you expect any change in that?

I would not say more recently. I think it's more of the same. The demand environment remains very similar to what we articulated on the earnings call. Only my hope is that as some of the impact of communications sector plateaus this year and stabilizes, you should see a slightly better outcome on CMT in coming quarters.

So I do see that there is an opportunity out there on the CMT space better than what it has been in the past. It's not because the environment is changing, but a particular customer that impacted our performance in the first half is now more stable and will not create any more negative headwind for us.

The question that comes up from investors is the, is that customer that created that impact, could that be more like that to come? Could there be other customers in CMT that make similar decisions in their spend?

Yes. I mean you can't say what would happen, but I don't see any of that on the horizon as we speak.

Okay. The other thing that we're hearing in the channel is that incumbents are more at risk than usual to be replaced by competing IT service companies. How much are the new kind of non-FTE models driven by AI impacting competitive decisions making incumbents more vulnerable?

I mean what I observed on the deal activity, including the large deals that Cognizant has signed over previous few quarters, is not whether it's incumbent versus a new player or a player from a new industry or something like that. I think the key differentiators have been your ability to demonstrate that you understand the use case or a large application of new technology like AI in a particular situation or customer problem.

Customers also look at which are the companies that seem to be the companies which are leaning forward and will be the companies of relevance in time to come, in the next 3 or 4 years. So combination of your ability to problem-solve a particular customer challenge and your ability to demonstrate that you would be that forward-leaning organization not just now, but in coming years, is the sort of secret sauce of winning the large deals.

Is there a pressure on incumbents? Yes. But it has always been that incumbents would always be challenged with an aggressive proposal from an outsider. At the same time, incumbents have the advantage of understanding the IT estate, which is more and more relevant. Most of the probabilistic solutions are AI plus context is equal to your answer. And ability of an incumbent to provide context is real, and it's a real advantage. So I would think it's not -- incumbents are not that much at a risk as you hear in the sideline conversations.

Okay. That's helpful. The other one that we discuss a lot on the industry is just the delivery model. And it's a surprise to us that we're still seeing headcount growth in some of the models. You would think that with AI, you would replace a meaningful amount of heads in that there would be -- there should be a reduction in the amount of heads for IT service companies of 25% to 50% or more like over a certain amount of years. So how do you see the headcount evolving in the industry?

Sure. So we track this quite closely. In fact, we have seen that from '23 to '24, there was an aggregate increase in employment or increase in workforce for large players; '24 to '25 also. But if you see '25 to '26, that number is largely flattish. And that is flattish despite most of us hiring a large amount of recent college graduates, and including Cognizant which continues to hire in a very large quantum.

So we are at a position of that transition bend that you see -- you no longer see that addition. You are still not seeing a reduction, but you are not seeing that addition. And that's where I think the model will be for next 18 to 24 months, where you will see more range-bound numbers around the current mean versus large additions or reductions.

But you should definitely see a more -- from a P&L standpoint, if I look at my cost of sales, I see my employee cost as the largest component now. Over a period of time, that should go down and should be replaced by the virtual effort or the inference cost. And still, on aggregate, not exceed today's total so that I am able to maintain my gross margins as we go.

Yes. And so over time, that headcount number, just in the industry, forget about Cognizant for a sec, probably at least starts to decline as the efficiency gains continue.

Yes. Continue to come.

Right. Got it. Another question we get a lot and we're getting asked is just thinking about the service budget, the allocation of dollars going to services. Is that getting squeezed versus other areas like tokens or memory getting a higher percentage of the overall budget?

Yes. So I think it is fascinating because it always works in cycles, right? I mean almost always when a new component in the ecosystem of outcome creation gets added, that new element has a disproportionate share of the total dollars.

I mean if you go back in history of when the first time licenses were sold by the likes of SAP, Oracle or Microsoft 20 years back, or if you go back and check when the cloud was being pursued aggressively in 2015, '16, '17 by Microsoft, Google and other large players, rightly so, that cost always increased as part of the total pie of the consumption for a CIO.

But as the technology became -- started diffusing, that tended to get spread over a period of time and it gave the space to the services. Because the value realization layer is the services layer where the customer sort of sees the outcome of that investment that it makes and that the customer has made initially.

So I feel very confident that this is a phase in the cycle where you would see a very high consumption of the new element, which in this case happens to be the tokens from LLM providers or the usage of GPU as it gets deployed. But as you see the value diffusion cycle, you would see that the spend sort of tends to balance itself out. Or it will balance itself up.

Yes. The other kind of question we're getting in the industry is thinking about this line blurring between software and services and that they're going to start competing versus each other, both -- going both sides. Service companies buying more product and more -- and then software companies getting into more services, implementation and maintenance. How do we -- how do you think about that lines blurring? And will the 2 sides start competing more aggressively for business going forward?

Yes. I mean it's a -- on the deterministic side, the boundaries are very clear where the SaaS sits and where IT services sits. When you go on the probabilistic side of the table and you see the role of the players, it is quite clear on the compute side, it's quite clear on hosting side, it's quite clear on LLM side. And downstream, the roles are evolving, and clearly, there is a little bit of blur.

I think over a period of time, that tends to stabilize and people -- and the companies figure out what they are best equipped to do and where they are able to add the maximum value for customer. Is there an opportunity for an IT services provider to be providing a bespoke AI solution to customers? The answer is a very solid yes. At the same time, for a SaaS player to do something that hitherto would have been characterized as services, probably the answer is equally a solid yes.

So it will evolve. I do see an opportunity for a blurring of the lines for next few years until a firm, clear line emerges here.

Okay. Great. Just thinking at a high level about Cognizant, you guys, I think, outlined at an Analyst Day a few years back that you guys wanted to get into the winning circle. And you guys have moved into the winning circle maybe even faster than you expected. And you guys are growing faster than other peers when it comes to organic revenue growth. Why is that? Why has Cognizant been in the winning circle faster than maybe others?

Yes. So thank you, and that was our aspiration, that by '27, we land that. But I'm very happy we landed that winner's circle position, we were able to do that in 2025 itself. And as I see the performance of first 6 months of 2026, we continue to be there at winner's circle.

And a few things have come together very well for Cognizant. One is that Cognizant always has been at the cross-section of functional knowledge, deep domain capabilities and technical sort of progress. And we see in the new world of probabilistic system or AI, that is a key advantage that is playing to our -- that is playing for us in any client situation.

Second is our ability to win large deals and then execute on them well, because winning a large deal is a virtuous cycle until you can continue to deliver very well on large deals. Because every large deal comes with a few very senior, very strong referrals that you are able to land the $200 million or $500 million deals. And those references are really somebody that you are serving today and serving the customers with a great satisfaction of the customers that those referrals come. So I think that's the second point, where we have been able to keep the virtuous cycle of large deal win and then delivery and then again winning them, et cetera, work well.

And third is I think we are investing disproportionately, relatively speaking within IT services industry, on AI. And customer sees us as a player of future. And therefore, you tend to win some of these large deals because customers don't only think about today, but 3 years from now when customers are deciding on meaningful customer projects. So I think these are the things which are coming out well. It's a good cycle of strong execution that we have been able to deliver to get to -- in that winner's circle.

Yes, I was going to ask about your guys' AI capabilities or AI strategy. How do you differentiate your AI strategy versus your competitors and your peers?

Yes. For longest time, I think IT services was less differentiated, if I can say that way. Because if you see between, let's say, 2002, 2026, all the new opportunities in the sector was offering-led. You were selling X, you could sell Y, you could sell Z, you could sell something new.

It was not as transformational as AI has been, which is not only selling new things but also changing the way you do your business. They're changing the way you perform your client obligations. You are now imagining almost 2 different industry or different stream lanes -- swim lanes of delivery: the one that was deterministic system and second is the probabilistic system. And we envision that all of our large customers in next 3 years will have 70%, 75% of the classic deterministic system because they will be still relevant, and another 20%, 25%, 30% of that, probabilistic system that they deploy.

Now this is a big shift. And we believe we have been able to differentiate because we invested before anybody. We remain almost very focused or very, very watchful of the fact that we are not only changing the front end and offering, but also we are changing the organization of Cognizant. And that's why in AI forum that we hosted, we had 1 conversation on our offering, but we had 6 conversations on how we are delivering that and what our customers are seeing.

So I think this is the first opportunity for one of the industry players to truly differentiate itself, by not only changing the offerings, which has happened for last 20 years in various scenarios, but changing itself. And that's what we believe is future. And that's what I think is also resonating with our customers. And therefore, you see the sense of differentiation that we are able to create with our customers.

Got it. So looking at bookings, you're up 5% over the trailing 12 months, led by growth in those large deals that you were referencing. Why is Cognizant able to -- I mean we're talking about this, I mean, obviously, it's AI led. But why is Cognizant gaining in the large deal wins? And is volume still growing double digits for you guys?

So I would say volumes is growing in the new work definitely double digit. But it is definitely -- in the existing piece of work, you definitely have what I would say productivity-led shrinkage. And therefore, on net, you are still growing. You are not growing double digits, but volume is still growing.

If I look at quarter 2 of 2026 over quarter 2 of 2025, there is a volume expansion. But it is a net effect. It is not a gross effect. Gross effect is double digit for new work, but it is definitely a shrinkage for something like software engineering, where we have ourselves said that 40% of the work is now AI assisted on software engineering side, so where there is a shrinkage in the human effort of work.

Got it. How much revenue today -- Cognizant has outlined the vector 1, vector 2 and vector 3 services. Just curious on -- because I feel like we're still heavily in vector 1. But how much are we seeing in vector 2 and vector 3? And how much has that changed yet '25 to '26? Or is that more a '27 to '28 phenomenon that we see vector 2, vector 3 maybe have a bigger impact?

You're right, we would see more impact of it in '27 and '28. On '26, we still see a similar pattern as we saw in '25, where the revenue is still dominated by vector 1 largely as proportion of revenue. But we do see a differing and more visible impact in bookings, which is an early indication that vector 2, vector 3 should eventually start contributing to a relatively larger share of the total revenue.

So in some of the signings that you're doing, you're seeing vector 2, vector 3.

Vector 3 opportunities.

And it takes how long before that converts into revenue typically?

Yes. It essentially converts within next 6 to 8 months. But they're still smaller deals. The pendulum of larger deals is still tilted largely on the vector 1 opportunities. So while they will start translating into revenue, by the time they start making an impact on proportion of revenue, it would be '27, '28. And hence, my comment that it's a little more out in the future opportunity for that revenue mix to shift. But we are already seeing visible clear wins which are no longer POCs, or $2 million, $3 million contracts. They are sizably larger contracts, but they are still not a $300 million contract.

So we have greater visibility into booking. They are more prominent features of our bookings now. But it is not yet shifting percentage of bookings or percentage of revenue yet.

Okay. Got it. Revenue per head increased 5% for Cognizant. Where can this figure kind of go to when we think about the revenue per head?

I think as we move more and more effort from -- towards inference and virtual effort, that number will continue to grow. I think this 5% to 7% to 8%, the range that we have spoken about in last few quarters, is a good benchmark to have. Because it's not going to be an overnight large shift. It's going to be a gradual shift as we generate more and more revenue through inference compared to the traditional efforts.

One of the conversations you and I have had is talking about the rate cards. And rate cards, obviously, in the beginning, were under pressure due to just the productivity gains that we talked about in vector 1. But there was a transition maybe in the beginning of this year where you started to see some supplement to the rate card on AI, getting a little bit of boost there and that obviously being a positive sign. Can you talk a little bit about the rate card, kind of where it was in '25 and maybe '26, and then how that might evolve with AI on top?

So I would say there is no specific pressure on rate card, as you rightly indicated, in 2026. It is more on the total cost of ownership. And as you look at total cost of ownership and fixed price still, it's very easy to compute saying, "You took all the risk and you delivered something at $100, now you can deliver at $80." I don't know and I don't care how much effort you put, both the classical effort and virtual effort, to get to that throughput. So there is no negotiation around individual's rate card in that sense on fixed price.

On time and materials side, we are increasingly going towards an A0 to A4 model where, on A0, we give price for a classical time and material rate card. And let's say, A1 is the primary doer is an AI and it's been initiated and reviewed by a human agent. Let's say A2 is something where primary doer is AI and only reviewed once before its submission by AI (sic) [ human ]. So there is a whole spectrum of A0 to A3 or A4, where A4 is a fully agentic system.

And depending upon the work that you perform, for example, application maintenance can go to A3 very quickly, whereas systems engineering or embedded engineering will be -- will probably never go to A3, it will always remain at A1 or A0. That's how the model is emerging on rate cards, not just for small deals, but very, very large commercial constructs.

So we are already embedding inference as part of our pricing. So on one hand, you could worry that the human effort is coming down and, therefore, P into Q, the Q is shrinking. But now you have Q1 and Q2 where Q1 is the effort which is classic human effort, and you have Q2 which is the inference, and you are pricing both in your T&M offering. So you have, of course, the Q1 is deflated, but Q2, which you never priced, so there is an opportunity for growth there.

Yes. No, it's interesting how it's evolving. I wanted to ask about Cognizant's expectations for organic revenue growth. I think, if I remember correctly, this third quarter is supposed to be a similar kind of organic growth as the second, or thereabouts. And then there is an implied kind of acceleration in the fourth quarter. Can you talk a little bit about the third quarter guide comparatively to the second? And then what's driving that acceleration in revenue growth for the fourth quarter?

Yes. So there are a couple of things. One, there is a slightly different days impact this year compared to -- working days impact this year compared to last year. So we do see a slightly higher number of billed days in the second half versus first half.

Two, and more importantly for us, is there have been a host of projects which have been on a transition phase during the first half, which we won in the last part of -- I mean, in Q4 of last year, Q1 of this year, that translates into the revenue addition in second half, which we have visibility to. I think so these are the, I would say, 2 reasons why we see the Q3 and Q4 guide numbers the way we have given.

Got it. And then can you talk a little bit about Project Leap and the impact to gross margins and the second half operating margin outlook as a result of that? And then as we get into fiscal year '27, how do we think about gross and operating margins?

Yes. So as we have called out, gross margin will slowly continue to improve as we go. But what we really guide on is operating margin. And we have guided, as a result of Leap, 10 basis points higher operating margin at the midpoint of the guidance range for 2026.

We are investing -- I mean, Leap is an exercise whereby we are creating a bucket of savings. And we are reinvesting a large bucket of that savings into employee training infrastructure, AI infrastructure, for delivery, investing in things like harness that we have built, and so on and so forth.

So to that extent, the savings of the Leap will be reinvested for an accelerated growth or staying ahead than others in the growth cycle. And we will make a decision of '27 as it comes. But for this year, we have articulated how that Leap saving flows into the operating margin cycle.

Okay. How about capital allocation? We've seen some more aggressive buybacks in the industry since the whole industry is down. Are you guys thinking about more aggressive capital return for buybacks versus what does the M&A pipeline look like?

I think we will always remain opportunistic on M&A space. This is also the time of transformation and time to invest in right assets. Of course, but one would be thoughtful about the fact that time -- this is also time to invest back in your own stock through buyback and other initiatives. So we have accelerated that in the beginning of this year. We also had an additional $1 billion that we announced and executed in the month of May.

Going forward, we will stick with our classic 50-25-25, 50% for M&A, 25% for dividend and 25% of buyback. And probably some of the future, the way to see it is that we will maintain this over a period of time, which means some of the future buybacks we have pulled forward in the current year because the timing and the price at which the share was trading that time, it was very opportunistic to do so.

Okay. We got a couple of minutes left. I have to ask about the latest on the Indian listing, how the mechanics of the listing will work. I know there's a lot of different things you guys are working with the regulators. So for example, how the IDR is classified. It's currently a derivative and limits the institutional investment, the limits on the use of proceeds, the tax ambiguity and the materiality of disclosure requirements, so all of those things. Can you give us an update on those conversations with the regulators and where we are on that potential listing?

Yes, absolutely. So I think there are various aspects around this. I wouldn't comment on one versus the other because, finally, it's a package of a regulatory framework that one would receive. And our expectation is that we should have some draft regulation on this being offered for players like us by the end of the year by the regulator. Once the draft regulation is available, Board will make a decision on whether the regulation is amenable from keeping an interest of our existing shareholders, keeping an interest of other stakeholders from a Cognizant standpoint, whether it makes sense.

Assuming it makes sense, then we go ahead and we tell the shareholders about it. If we decide not to, then we share accordingly with the shareholders. So I think we will have some decision on it by end of this year, is what our current anticipation is.

And my guess is there's some give or take, or like some of the things that you would want, you definitely want to get in there or want in there, you're not going to exactly get. And then the regulator is going to have to change some things that maybe they want to change. How are those dialogues going between -- I assume both sides want to try to get something done, but is it even feasible given some of the complications?

Yes. I think the dialogue has been very constructive, and we are very grateful for the conversations that we have had with regulator. And one hopes that finally we are able to find a ground which is a great product for, not just Cognizant, but any other company which wants to pursue, that there's a great product out there for everybody to pursue. But right now, that's all I can say because we will have to wait for what comes out in the final regulation.

Yes. Okay. With that, Jatin, we're going to have to keep it there. Thanks so much for being here.

Thank you very much, Bryan. Thank you.

Ladies and gentlemen, welcome to the Cognizant Technology Solutions Second Quarter 2026 Earnings Conference Call. [Operator Instructions]. I would now like to turn the conference over to Mr. Tyler Scott, Senior Vice President, Investor Relations. Please go ahead.

Thank you, operator, and good morning, everyone. Welcome to Cognizant's Second Quarter 2026 Earnings Call. I am joined today by Ravi Kumar, Chief Executive Officer; and Jatin Dalal, our Chief Financial Officer.

By now, you should have received a copy of the earnings release and investor supplement. If you have not, copies are available on our website, cognizant.com. Before we begin, I would like to remind you that some of the comments made on today's call and some of the responses to your questions may contain forward-looking statements. These statements are subject to the risks and uncertainties as described in the company's earnings release and other filings with the SEC.

Additionally, during our call today, we will reference certain non-GAAP financial measures that we believe provide useful information for our investors. Reconciliations of non-GAAP financial measures where appropriate to the corresponding GAAP measures can be found in the company's earnings release and other filings with the SEC.

With that, over to you, Ravi.

Thank you, Tyler. Good morning, everyone. Thank you for joining us. We delivered a solid second quarter with organic revenue growth at the high end of our expectations and year-over-year adjusted operating margin expansion. Nearly all our healthy sequential growth was driven by our organic business. We accelerated our evolution as an AI builder by building new capabilities, launching new platforms and deploying frontier talent as we begin to unlock entirely new business categories and client value pools.

Looking at the quarter's highlights. Revenue grew 4.1% year-over-year in constant currency, led by strong performance in North America as large deals signed over the past year moved into full execution. Financial Services grew nearly 12% year-over-year in constant currency, its second consecutive quarter of 10-plus percent growth. Trailing 12 months bookings increased 5%. We signed 7 large deals, each with TCV of more than $100 million, including 3 new logos. As we expanded adjusted operating margins year-over-year for the sixth straight quarter, demonstrating continued profitable revenue growth.

From an AI indicators perspective, our revenue and adjusted operating income per associate increased 4.6% and 7.1%, respectively. Starting this quarter, we are excluding trainees who are not fully deployed for both the current and the comparable prior periods. Over 40% of our software development is now AI-assisted. We have over 8,000 AI engagements. And we view strength in Financial Services, which include some of the world's most technically sophisticated companies as a leading indicator for other industries. Our research reveals that financial services is well ahead with AI initiatives and advanced AI adoption.

We are helping clients tackle significant technology debt by using AI to compress modernization time lines that are shifting towards outcome-based pricing. We also now see financial services clients leveraging AI for growth imperatives with new discretionary spend cycles. We completed our previously announced acquisition of Astreya, a global IT managed services provider with deep expertise in data center infrastructure, enterprise networks, digital workplace services and AI first managed operations. Momentum is already building. Astreya will continue its work with Google to deliver services across its corporate engineering environment, including global ID Ops, workplace collaboration infrastructure and platform services.

We delivered these results against the cautious demand environment while growing at the top of our peer group. While we expect that caution to persist in the near term, AI is driving fundamental change in our industry, that we believe creates a significant long-term growth opportunities. The key question is why growing AI capability has not yet translated into more enterprise value. Our research shows 2/3 of the Global 2000 have not yet realized measurable AI productivity gains, 1 in 4 have paused AI deployments and billions of dollars in potential value remain unrealized.

The opportunity to address this gap is enormous. We estimate the $1 trillion system integration market can expand into $5 billion to $6 trillion enterprise operations market with $4.5 trillion of operational labor exposed to AI. Services firms are structurally positioned to capture this opportunity. As models proliferate and inference costs decline, models stop being the differentiator and value shifts to the applied layer, which is context governance and business processes. This layer addresses how each enterprise applies AI, what it learns from it, how effectively it retains, reuses and compounds that learning and how well it protects the proprietary intelligence that constitutes its alpha.

That is why we launched the Cognizant AI Delivery Operating System, a continuously learning delivery system that combines human expertise, organizational knowledge, client context and AI intelligence. It has 3 pillars: our engineering harness, which causes engineers in real time; our business operations harness, which feeds best practices into a shared organizational system; and our intelligent Spine, which connects intelligence across physical and edge environments. All of this is supported by our context engineering capabilities, which is the ability to assemble and enterprises work graphs, guardrails and tribal knowledge.

Working with Cisco, we are using Context fabrics to build a digital nature of the account management role forming the foundation for an account managers digital twin and broader identification solutions. This AI solution with a digital twin at its core holds the potential to streamline daily operations and support on time in full complete deliveries, significantly improving customer satisfaction. For a large North American bank, we piloted a solution for their fraud dispute management and KYC operational workflows. Context was engineered through a custom solution that combines static knowledge from customer and operations interactions and documents with dynamic business content sourced through enterprise application APIs.

The fraud dispute management, multi-agent pilot solution has demonstrated the potential to reduce manual effort by more than 50%, while the KYC process can significantly improve decision efficacy, reducing the risk of fines and fees. Last quarter, I described how AI is reforcing our industry's first principles and driving 4 significant shifts. First, becoming an AI builder rather than a traditional systems integrator, owning the full stack required to design bespoke AI systems. Second, rebuilding a talent model by shifting from a traditional pyramid towards interdisciplinary teams working at the intersection of domain operations and technology. Third, shifting our economics from labor to outcomes. Our mix of fixed price and transaction-based work has grown for 3 consecutive years, creating a more durable business. And fourth, moving from delivering projects to underwriting results.

Let me share our progress across these 4 ships, starting with the first one, becoming an AI builder by strengthening our proprietary IP and ecosystem. This year, we launched a dedicated AI market unit, an elite team of business designers, industry strategies and frontier engineers, focusing on converting our investments into realized value. We're already seeing early traction. For example, a large payer client chose us to help build an agentic development practice for its biggest division through parts of frontier engineers and AI agents. We cut manual effort by 60% for a midsized payer while improving the claims throughput. And we compress AI development cycles from months to days for a leading European online fashion retailer, advancing agentic workflows across supply chain inventory returns and customer experience.

On the partnership front, we established a dedicated Gemini enterprise practice as a Google Cloud diamond partners, and we joined OpenAI's Daybreak consortium. We also expanded our partnership with Anthropic Begin, one of the small number of global premier partners in the cloud partner network. An example of this partnership at work is Travelport, where we partnered with Anthropic on a strategic AI transformation aimed at modernizing Travelport's software development and embedding AI across Travelport's travel retailing and distribution platforms.

And with A+E Global Media, we partnered with Snowflake to deploy custom intelligent agents that transform complex advertising operations and legal workflows. By automating document validation and enabling natural language queries, we helped accelerate decision-making and significantly improved time-intensive processes. For our second shift, we are rearchitecting our talent model into an AI augmented early career talent led by more senior player coaches. We introduced 2 new certified roles, frontier certified engineers who audit workflows and build intelligent reasons and Frontier business operators who manage blended human digital teams to deliver outcomes.

We plan to scale this cognizant forward team to 5,000 frontier certified engineers and 10,000 frontier business operators. We currently have 10,000 cloud-certified architects, the most of any organization globally, and we power one of the largest pools of Codex and Gemini enterprise trained badges. Last quarter, we introduced Cognizant SkillSpring, an AI-native platform that embeds agent-driven tutoring directly into daily workflows and gives associates real-time visibility into their AI proficiency and token usage. It is gaining significant momentum with our associates as learning time has doubled and AI uses has tripled.

We have also opened this platform to early prospective clients. Our third and fourth shifts move Cognizant from a labor-based model to an agentic and platform enablement model and from delivering outcomes to underwriting results. This is why we established a new AI products and Platform Group earlier this year to unify Cognizant's proprietary offerings and scale innovation across the portfolio. Our platform strategy has 2 dimensions: first, our engineering platforms, which provide the foundation for everything we built, including accelerators, agent frameworks and AI engineering tools that power our AI native software development life cycle and agent development life cycle. They're increasingly powered by their own Agentic workforce. Together, they compress the software cycle improve productivity and accelerate business outcomes for clients.

Second, and building on that foundation, our business platforms combine technology, data, AI and deep industry expertise to create differentiated client value purpose built for the industries we serve, the embed industry-trained agents directly into critical workflows. TriZetto is the strongest proof point of our platform strategy. Our health care platform business generates more than $1.1 billion in annual revenue. And through the first half of 2026 grew faster than the overall company while delivering substantially higher margins. What began as a software product has evolved into a broad health care platform ecosystem.

TriZetto demonstrates how platforms can drive deep client relationships, create recurring revenue streams and deliver growth and profitability that exceeds traditional services. It's a blueprint for how we intend to scale platform led growth across other industries. In health care claims, we built a pioneering auto adjudication solution that uses large language models to digitize adjudication knowledge and rules, enabling Agentic AI to analyze claims, apply complex business rules and reach decisions with human validation wherever it's needed. It positions us to take a share in this large, high-volume category.

Other platform-led modernization wins include a global claims and risk administration leader, deviating neuro AI flow source and our 3 cloud-enhanced Microsoft expertise, we signed a 5-year agreement to accelerate processing times and upgrade core infrastructure. And Cotality, a global data and analytics company has deployed artisan's neuro business process workflow across multiple business operations processes. It resulted in over 40% improvement in research as turnaround times and faster and higher quality resolution of their customers.

We're also moving beyond delivery to underwriting results. We signed a major engagement with a leading insurance brokerage committing to more than 50% productivity improvement over 5 years through AI and operating model redesign. We won on the strength of a domain expertise, reimagining the core workflows and accelerated delivery using AI tools from our partner ecosystem. As we execute these 4 shifts, our AI builder model expands where we create value across 3 categories. First, our traditional work done dramatically more productively. Second, old things in new ways and third, entirely new things that didn't exist before AI. First, our traditional work done more productively. This includes autonomous software engineering or vector 1 work, which over the past 2 years has driven both consolidation and productivity-led engagements.

A great example of our success in this area is Novartis, which selected Cognizant earlier this year for a 5-year engagement to transform its global ID operations. Building on a relationship that spans more than 20 years, we expect to leverage our neuro AI platform to create a unified AI-powered operating model that combines automation, full stack observability and agentic capabilities. This is where our AI builder strategy is aimed at helping clients move from labor-intensive operations to intelligent self-service and increasingly autonomous technology environments. Second, all things in new ways. Here see a significant pipeline across secure AI services, mainframe monetization, SAPS for HANA migration and SaaS reimagination, long-standing enterprise challenges that AI can now solve far faster refraction of the cost.

Cybersecurity is a clear example. AI is turning security from a cost center into a remediation opportunity as machines expose vulnerabilities at unprecedented speed. We are positioned for this moment by combining frontier models with a 5,000-person security practice and all the leading frontier program partners, including CrowdStrike, [indiscernible] networks, Zscaler, Anthropic, OpenAI, Microsoft and Red Hat. We see a strong pipeline forming across these partnerships.

Third, entirely new things made possible by AI, including context engineering, reinvention of business flows, identification of business operations and physical AI. With physical AI as intelligence begins to govern physical environments, we believe a new domain of autonomous operations will open. We launched our solver in physical AI platform as a service to position Cognizant ahead of this iPhone moment for robotics and infrastructure. This builds directly on a capability we've built over the past decade. More than 10,000 of our associates have trained AI and machine learning models for the world's largest technology companies.

We are now repurposing that expertise for the enterprise through our AI model training and data services. For example, for a global automotive manufacturer, we train models on the company's products, technical data and visual content to achieve accuracy generic models cannot match, automate complex processes and unlock value from knowledge the company already owned. Public sector is emerging as a meaningful business as our organic and inorganic investments gained traction. In Cognizant government solutions built on our Belchan acquisition secured a landmark engagement with the state of Iowa to modernize its IT infrastructure.

We have a growing pipeline across defense, federal and state government, including AI infrastructure and Citizen experience, while TriZetto expands the opportunity to health agencies, including our work supporting the Department of Veteran Affairs in partnership with Signature Performance. This builds on our long-lasting U.K. public sector practice, for example, with the home office, we develop and support the foundational data platform behind the U.K.'s migration and border systems. In Q2, we won expanded home office work across software engineering, testing, delivery and managed services for critical case working systems, improving case worker productivity and reducing manual intervention.

And for his Majesty's revenue and customs, we recently won an additional work to help configure low-code services in support of build and DevOps functions that is valued at more than $215 million over the life of the deal, including optionees. Our AI Builder strategy is also gaining traction outside the U.S. For example, a global pharma company in Europe selected Cognizant as a sole partner to build and scale its enterprise data, AI and agentic AI capability through a 3-year agreement covering 68 projects initially as the clients official builder Cognizant will translate their agentic KI vision into a production-grade governed enterprise platform spanning all business domains globally.

And Cognizant helped a large European bank to identify its mortgage processes by building a mortgage operations agent, which brings multiple specialized agents together to support complex decision-making, analyze business rule outcomes, proposing remediation paths and generating clear and actionable insights.

To conclude, we're in the midst of a profound transformation with a clear vision for the industry's future and confidence in the expansive AI-led opportunity ahead. Our actions, deploying interdisciplinary talent, shifting to outcome-based platforms and opening new value pools are designed to drive sustainable growth. As we redefine Cognizant, we remain focused on our growth on our goals of delivering top-tier growth, consistent margin expansion and EPS growth ahead of revenue. Thank you to our associates, clients and shareholders for your continued dedication, partnership and trust.

With that, I'll turn the call over to Jatin.

Thank you, Ravi, and thank you all for joining us. We are pleased with our second quarter performance, highlighted by industry-leading growth and steady adjusted operating margin expansion. We achieved these results while continuing to invest, including in the completed acquisition of Astreya, new frontier skilling initiatives, expanded partnership and our AI labs and platform-led offerings. We also deployed more than $1.1 billion through share repurchases, reflecting our conviction in the long-term opportunity AI creates for Cognizant and our critical role as an AI builder.

While market conditions remain complex, we have continued to deliver on our commitments while investing in and evolving our business for the future. Now moving on to the details of the quarter. In Q2, revenue grew 4.1% year-over-year in constant currency to $5.5 billion. Our sequential organic growth was at the high end of our expectations. Year-over-year performance was driven by volume growth increase in third-party product revenue associated with our integrated offering strategy and inorganic revenue from our investment in 3Cloud.

From a geographic perspective, growth was once again driven by North America. And from a services perspective, our BPO practice once again led growth while demand remained strong for data and cybersecurity driven by AI adoption. We are also seeing strong growth from industry-specific AI-led transformation in Financial Services and Life Sciences. By segment, Financial Services again led with healthy growth across banking, capital markets and insurance clients. Growth is also being driven by strong performance in the U.K. public sector. We are seeing legacy modernization programs accelerate as clients advance their AI journeys to address significant technology debt. This is also reflected in sustained bookings momentum.

Health Sciences was stable. Demand remains cautious and cost-driven with clients prioritizing vendor consolidation, legacy modernization and compliance while discretionary spend faces tight scrutiny. It must demonstrate a clear ROI. As Ravi mentioned, TriZetto had a strong quarter. Products and Resources was steady. While clients in retail, consumer goods and travel and hospitality continue to navigate pressure from geopolitical uncertainty, supply chain disruptions and elevated oil prices, we are seeing momentum in manufacturing, logistics, energy and utilities, where physical AI and smart manufacturing are creating compelling opportunities for us.

Within Communications, Media and Technology, demand among comps and media customers is muted, consistent with last quarter. With technology customers, demand remained strong, driven by AI native engineering, digital operations, data and cloud services. As Ravi noted, we are already seeing momentum with Astreya and we are confident our joint capabilities can generate attractive growth synergies in the years ahead, driven by AI infrastructure buildout.

Turning to bookings. We delivered another strong quarter of large deal bookings, signing 7 deals each with TCV of more than $100 million, including 3 new logos. On a trailing 12-month basis, bookings grew 5% and represented a book-to-bill of 1.3. Annual contract value decreased modestly, reflecting the impact of lengthening contract duration due to a greater mix of large days. We are pleased with the growth we have delivered in new and expansion bookings, which grew in the mid-teens in the first half of the year.

Moving to margins. During the quarter, we incurred approximately $84 million in cost related to the project lead program we announced last quarter. In addition, as a result of India Labor Code and subsequent regulations notified by the Indian government in Q2, we recorded $81 million onetime benefit for a partial reversal of the India defined contribution obligation liability we had originally recorded in 2019. Excluding these impacts, second quarter adjusted operating margin was 16%, up 40 basis points year-over-year. Operational efficiency and favorable currency movements more than offset higher third-party costs and compensation costs as well as the impact of our recently completed acquisitions.

Now to details of EPS, cash flow and capital allocation. Second quarter adjusted EPS was $1.37, up 5% year-over-year, driven by revenue growth, margin expansion and lower share count. EPS was negatively impacted by a higher interest expense associated with $1 billion we borrowed under our revolving credit facility to fund the Astreya acquisition and share purchase activity in the quarter. DSO was 88 days, up 5 days year-over-year, primarily driven by a change in the business mix. This factor also led to a corresponding increase in payables. And therefore, the impact was neutral to cash flow.

Second quarter free cash flow was $459 million, bringing year-to-date free cash flow to $652 million. During the second quarter, we deployed $1.1 billion on share repurchases and bought back over 22 million shares at an average price of approximately $51 per share. This includes $500 million accelerated share repurchase program announced in May. Year-to-date, we have returned $1.9 billion to shareholders through share repurchases and dividends and remain on pace to return about $2.6 billion. This represents more than 10% of our current market cap. We have also deployed $1.3 billion on acquisitions aligned with our AI builder strategy. Finally, we ended the quarter with cash and short-term investments of $1.1 billion.

Turning to guidance. For the third quarter, we expect revenue to grow 3.8% and to 5.3% year-over-year in constant currency. This includes approximately 200 basis points from our recently completed acquisitions. As we discussed on our last earnings call, our prior guidance range contemplated and improved discretionary spending environment at the midpoint. Instead, macro uncertainty has remained elevated. We have therefore revised our full year revenue guidance range to 4% to 5.5% growth in constant currency. This includes 150 basis points of inorganic growth unchanged from our prior expectations, but similar to last quarter, our M&A pipeline remains active, and we are focused on executing with discipline on opportunities aligned with our AI builder strategy.

While discretionary spending has remained pressured, we have maintained good traction on large deals, which we will expect to continue to ramp in the back half of the year. Our revised guidance range assumes the discretionary spending environment remains stable at the midpoint while the high end contemplates an improvement in short-cycle revenue in the fourth quarter. There are no changes to our project lead cost estimates or expected savings, and we continue to expect the program will run through the remainder of the year. Our adjusted operating margin guidance is unchanged at 16% to 16.2%, representing 20 to 40 basis points of year-over-year expansion. Our free cash flow conversion guidance for the year remains 90% to 100% of net income.

Full year tax rate is now expected to be towards the low end of our prior guidance range of 25% to 26%. Based on our current expectations, we expect our third quarter rate to be above the high end of the full year range. We now expect full year weighted average diluted share count of approximately 460 million down from our prior estimates due to the pace of repurchases in Q2. Interest expense has also increased modestly, reflecting a lower cash balance and the drawdown of our revolver this quarter. As a reminder, the previously disclosed enactment of the Indian Labor Code reforms in 2025 has resulted in a higher run rate of other expenses.

We expect this below-the-line costs related to our India defined benefit plan will be around $10 million per quarter for the foreseeable future, consistent with the first half 2026 run rate. This is in line with our estimates in our initial guidance in February, but we are highlighting it to support your modeling. Our EPS guidance has increased to $5.70 to $5.82, representing 8% to 10% growth versus 7% to 9% growth previously.

Finally, we continue to make progress and advance on our evaluation of a potential primary offering and secondary listing in India. We are working in close collaboration with external stakeholders and regulators. We will make a decision on this once we have visibility of the revised regulatory framework. We are pleased with the progress made to date and remain committed to acting in the best interest of our shareholders. We will provide updates as appropriate.

With that, we will open the call for your questions.

[Operator Instructions]

Our first question today is coming from Maggie Nolan of William Blair.

I'm hoping you can give us a little bit more commentary on the business momentum in the context of bookings growth compared to last quarter as well as that second half ramp-up that you had previously expected from large deals. Maybe update us on how those signings and ramps are progressing and how it now shapes your second half expectations.

We continue to have good bookings momentum. Last quarter, we did 22% bookings growth. TTM this quarter has been 5%. If you take the first half, it is 6%. It's a tough compare also because we had 2 mega deals last year. And last year, we grew by almost 18% in quarter 2 last year. So keeping all this in context, I think we have done pretty well on bookings, and I actually feel very confident about bookings for the rest of the year as well. Now one of the nuances which we are excited about in our bookings momentum is Financial Services is really running hard. I mean you've seen in Q4, we had 9% growth, 9-plus percent. In Q1, we had 10-plus percent and now 12%. So Financial Services has literally -- has overwhelming increase in bookings in the first half, and I expect that to remain very strong in the second half.

With its 7 large deals, 3 new logos, we are starting to see activation of $50 million to $100 million deals, which have significantly improved. I mean, if you take those 2 large deals out and compare from last year, $50 million to $100 million deals have gone through a massive bump. $25 million to $50 million deals have gone through a massive bump. Percentage of new business has bumped up by 10% in the first half in comparison -- 10-plus percent in comparison to the mix, which is also good because it kind of translates to new revenue -- incrementally new revenue for the second half.

So we've had a pretty good step-up change in our bookings momentum. When we entered the year in 2025, we were at $27 billion TTM, and we got to $28 billion TTM. And in the last 2 quarters, we are at $29 billion TTM. So we are starting to move up. And bookings are going to be a little bumpy between quarters. But if you look at the aggregate numbers and you look at TTM and you look at the tail velocity of the last 2 quarters, we feel super excited about the second half as well.

And then on the BPO business, you've seen good traction there. Can you talk a little bit more about where you're seeing that traction from an end market perspective? Is it really your vertical expertise that's helping there? Or is it more the partnerships that you outlined with some of the model providers and others that are important in this space? What's driving the success there? And how can you perpetuate it?

Great question. In fact, BPO has always been a blockbuster service line for Cognizant over the last 3 years. We continue to lead on industry vertical over the last few years. In fact, even when I came on board in 2023, the BPO organization was called intuitive operations, so it had embedded itself with data and automation and machine learning then and now AI-led BPO. Maggie, I've actually mentioned this in my remarks as well as in all my commentary in the last 1 year. The expansive opportunity of system integration services goes from a trillion-dollar market where we build software systems for companies to embedding technology, which is AI technology, agentic into business operations of firms. And that is going to move our market from $1 trillion to $5 billion to $6 trillion, and it is actually much, much more expansive than ever before, and it kind of brings data, technology and process altogether.

So we're very excited about the BPO business with the strength of model company partnerships where we can not just apply it for software engineering, but applied for business operations, vertical and horizontal, and also platformize that business. I mean our TriZetto business is running at a higher velocity than the rest of the company and the BPaaS business underneath it, which is health care operations, is equally running with the same velocity. So we want to replicate the platforms, AI-led agentic business operations for companies.

Just to give you an example, we have a blueprint for F&A, frontier F&A. We have a blueprint for frontier-led customer operations. So we've kind of started to put that in the mix, which effectively means we can embed digital labor and human labor and deliver outcomes through frontier operators as we call -- as we coined it. It's a new archetype of role and deliver those services to our clients. So I'm very upbeat about the future of our business process operations and agentic-led business process operations.

We also have a training capability now, which is the AI data training services. Historically, we did it for the big -- for the Magnificent 7 companies. Now we are transitioning that capability to AI-led into the Global 2000 because if intelligence is not going to be drawn centrally and if enterprises are going to build distributed intelligence, they don't have their own specialized models, we think we have a unique service to attach to it. We have 10,000-plus associates who work on data training services. That's a part of the BPO organization.

The next question is coming from Jim Schneider of Goldman Sachs.

Ravi, I think relative to your comments about corporate 1 out of 4 sort of pausing their AI progress because of cost or return issues, can you maybe talk about more tactically? I mean you've talked about how Cognizant can address that opportunity, but can you maybe talk about more tactically what customers are doing then? If they pause, what is their sort of immediate action? Are they going back to sort of more traditional implementation work or outsourcing work? Or are they just sort of pausing until they can get a better handle on the scenario? And how long would you think it would be on an average engagement before you can really see for Cognizant, a big uptick at customers like that?

Jim, great question again. Thank you. Look, the first chapter of AI adoption was broad-based, open-ended experimental, and this is a magical -- the technology was magical. So everybody tried to use it in a way that they could find some magic coming out of outcomes. As you productionize this, which is the second chapter, you're going to go very nuanced, you want to start to focus on not token consumption but token economics. And you're going to start to optimize where you use advanced reasoning and where you don't use advanced resin. So the step back from clients is to say, wait a minute, I'm spending a lot of money on on tokens. I'm spending a lot of money on the entire AI stack, Am I getting the value? And if I'm not, let me revisit how to optimize it and get value out of it.

The capabilities out there, I've said this, the production value is way below and the bridges to that production value is assembling context, assembling tribal knowledge, setting the guardrails, grounding the technology into the heterogenety of an enterprise. And we have started to believe now that we have a role to play, a big role player in that process, starting from building the harnesses, where you can capture the context so that when you do the transactions on a regular basis, you can create repeatability, model routing, which means depending on the kind of task, you could use an open weight model, you could use a costly, expensive, close frontier model or you could use a cheap or close France model or you may not use a model.

And then creating learning loops between human effort and machine effort, so that you could integrate human and machine for together, which means we have a methodology called basis in our consulting organization, which allows us to reinvent and reimagine those processes. So if you put all of this together, there's a lot of heavy lift needed before you can actually productionize it and get value. And we are building platforms and we're building services underneath it. Our clients are coming back to us for a variety of things, starting from productivity, which is related to software engineering, which was historically for the last 2 years, very mainstream.

Now going back to my previous response, they're coming back to us on business operations. I mean business operations is where the future is because you want to embed this technology into business operations if $6 trillion of AI ramp on the infrastructure is going to happen in the next 3 years, $15 billion to $20 trillion has to come out of value from enterprise businesses. And that's not going to come from system building through AI-first software engineering, but it will actually come from embedding it into business operations. So clients are using it for -- using us for getting the frontier capacity, the frontier engineering and frontier operator capacity, platforms, harnesses, context engineering, reimagining the workflows. And some of our clients are starting to ask us to deliver an AI-infused rate card, which means you embed the pre-training costs and you embed the inference costs into that process.

Software engineering is very mature business operations is actually evolving now. And we have a third hardness for physical AI, which we are preparing, which we think will be the future as we go forward. So that's the broad story. So the ability to build that bridge is what will drive companies like us to add value in the process.

And then maybe as a follow-up, financial questions sort of maybe for Jatin. Can you maybe talk broadly to sort of your overall hiring headcount plan in relation to gross margins. So it head count ticked down a little bit sequentially. I'm assuming a lot of that was just project LEAP and some efficiencies there. But maybe talk about your hiring plans over the next 2 or 3 quarters. And to what extent do you expect to be able to hold gross margins at or above the current level?

Sure, Jim. So as you rightly observed, we have flattish headcount between quarter 1 and quarter 2. We continue to add the recent college graduates to the company, as you indicated in the past, and we have made good progress by the end of first half, and we remain on track to get to approximately 20,000 by the end of the year. The project LEAP will is also underway. And as a result, you will see a certain amount of head count reduction. So on the balance, we expect that the head count should remain range bound versus an increase. We -- and that's how we are budgeting or we are planning for rest of the year.

So long as gross margin is concerned, you have, I'm sure, noticed the improvement that we were able to execute between quarter 1 and quarter 2, which is roughly 60 basis points. We are still trending a little lower than last year. And we will continue to work on it during the course of the year. And I do hope that we continue to show an improvement in that number as quarters progress.

Our next question is coming from Jamie Friedman of Susquehanna International.

Good results here. I wanted to ask about the linearity of the remainder of the year. And Jatin, the sequential assumption on the Q4. It looks like if you're at or just above the midpoint on the Q3, you could be flat to slightly down in the Q4 sequentially. But there is some M&A in there. So if you could help us think about how you're thinking about the sequential Q4 in particular, on an organic basis, that would be helpful.

Sure. So we have modeled it based on the trends that we see every year, what we have super imposed this year are a couple of variables. One is the larger new and expansion percentage of bookings that we have seen from the beginning of this year. We also have seen the ramp-up of the deals, which are in transition phase now and which will move to more to billable volumes in quarter 3, quarter 4. And finally, we do have a view on furloughs. As you know, the furloughs typically are represented largely by banking and financial services, and that is continuing to be very robust this year. So we have assumed a slightly lower proportion of furloughs coming in quarter 4. So the assumption is a slightly superior sequential growth in quarter 4 compared to what we have seen traditionally in quarter 4, which is typically negative because of the builders impact and further.

Perfect. And then, Ravi, I just want to ask about Products and Resource. It's been a couple of years now since Belcan closed. You had a ton of inorganic in the period of comparison in the Q3. So at a higher level, though, how is Products and Resources performing relative to what you had expected when you closed the deal?

Yes. So that's a great question. In fact, one of my endeavors is to go beyond Financial Services and health care and create more diversity in our portfolio. So we're very pleased with the performance in Products and Resources over the last few quarters. We've got some good traction, new logos. You've seen a section on my earnings around Belcan and its tailwind with other public sector opportunities. We have one using the Belcan engine. That is a great add to our portfolio mix. We are starting to see significant traction with our clients on physical AI, which I spoke about. We have built a harness around it. It's called the Intelligence Spine.

And we have just now hired a new leader for oil and gas. So we are continuing to make good progress on diversifying our portfolio and and Products and Resources is one of the important areas to do. I actually believe the AI opportunity will actually be in you will see a leap frog of digital enhancement on physical things in Products and Resources, and you will equally see low-margin businesses, which is what I mean by it is our clients, our enterprise clients, they're going to use AI to unlock more value than high-margin businesses just because of the productivity opportunity there. So Products and Resources is going to be one of our high investment zones in the future, and we'll continue to invest to make it a very important portfolio for Cognizant.

The next question is coming from Darrin Peller of Wolfe Research.

All right. Can you just touch on how you'd assess the competitive dynamics in the market right now, just especially for the larger deals, how important is pricing in these discussions right now? And then just when you when you're having these discussions with customers, what do they look like when large deals come up for renewal, just productivity savings they're demanding now versus prior?

Yes. So look, we have done productivity-led large deals over the last 3 years. We have outperformed on our margin performance versus what we originally assumed. We have -- we've done pretty well in winning more than 2.5 years. And that has led to a large deal momentum over the last 2 years. Now we have progressively gone into newer things, which is my second and third stream lane, which is doing old things in new ways, which is primarily, say, do a mainframe migration using frontier models or I do SAP HANA migrations or I do vulnerability remediation coming out of security or SaaS reimagination.

All those are starting to become large deals. And new things in new ways, which is primarily using AI to do things which we didn't do before for growth imperatives or smaller deals because they become more modular. So the mix of deals has changed as we progressed on this process. And on -- specifically on productivity, look, unlike in the past where you had circuit breakers and how much you could do on linear productivity because labor costs are not as nonlinear. Now you have a level of nonlinearity because you could pass on that productivity to clients and out what you have actually committed to clients and keep some for yourself. And that's why our margin profile on all our large deals, both $50 million a barrel and $100 million a barrel. It's actually trending much better than what we originally signed the contracts for.

So as long as we stay ahead on AI-led productivity for software engineering and business process operations, and we keep staying ahead of it, we can pass on the productivity, stay competitive in the market and still be margin accretive for ourselves. And that flywheel is working very nicely for us. And that's why we are continuing to win large deals, which are productivity-led. We will start to see that move from software engineering to business process operations where the span -- where the expansive opportunity is going to be much, much more.

All right. That's helpful. Maybe just a quick follow-up would be around what the path looks forward for scaling the frontier certified workforce that you described earlier. Just what degree will this come from new hires versus existing? And then just how are you going to keep differentiating as other companies try to develop front our workforces also?

We're doing it at scale. We are hiring at the bottom of the pyramid from outside and we're doing it at scale, building bridges from inside. Just look at where we are. Earlier this week, we announced we have the largest pool of cloud certified architects on the planet. We have 10,000 plus. We just finished the hackathon today with open AI in India, and we have 10,000 associates getting badges. We're doing a similar exercise with Gemini. We are also doing a lot of work with open weight models. So at scale, bending the cost curve and having the context of businesses to deploy the talent and get value out of it is what will drive companies like ours to be on the cutting edge.

You need a combination of things. You need to know how to reinvent the flows. You need to know how to audit the flows, integrate the agentic work into the business flows. You need to know the context and you need to do it at scale at a lower cost, bending the cost curve. And that's what we're doing. So I'm pretty confident that we will have the largest pool of certified frontier engineers and frontier operators, which is a new archetype of a role we have established. Frontier engineers is about engineering agentic tech into business flows. Fronter operators is about operating those flows, which has digital and human labor together. And doing -- bending the cost curve is what we have done for the last 30 years at scale, and that's what we are continuing to do. So it's a combination of building a pipe from outside, building a pipe at the bottom of the pyramid, early carriers and building bridges from inside.

The next question is coming from Tien-Tsin Huang of JPMorgan.

I just want to ask around Financial Services. That was up double digits. It's growing at a premium over the other sectors. In the past, we've looked at that sector as a maybe a leading indicator of another subsectors would follow. Do you see that potentially being the case here? Should we be encouraged that, that could be the case? Or is there something unique that maybe is a little bit different in terms of their willingness to to adopt some of these AI-driven projects?

Absolutely, Tien-Tsin. Thank you. Look, Financial Services has always been a pioneering industry. high on technology spend, they create a symmetry using technology, and they're on the cutting edge on AI. In fact, Financial Services, we have -- we are probably the #1 company on growth in our peer group. All of last year, we did higher single digits. At the end of quarter 4, we got to 9%. Quarter 1, we got to 10%. Quarter 2, we are now at 12%. And Financial Services is activated on all 3 swim lands, starting from consolidation, productivity, sharing the productivity to modernization of landscape using AI and refactoring landscape, which is my second swim land to the third swim land building new things using AI and generating growth imperative.

So I am actually super optimistic that financial services will lead the path and other industries will follow. And other industries will, sometimes, leapfrog as well. I mean I now start to see that in industrial clients who are looking at the miss they had in the digital revolution to leapfrog directly into physical AI. So you're absolutely right. I think it's a leading indicator to what's going to come. And if you look at our quarter 4 exit rate, just as a company, it's also powered by Financial Services. We are super excited about the fact that we are exiting -- if you take the midpoint of our guidance range, we are exiting on a high, and we will be on the winner circle.

Good. And then maybe for you, Ravi and Jatin. Just thinking about -- I have to ask question on economics, if you don't mind, just thinking about that, any update with respect to cost usage, what you're hearing from your client base. Any update there? I know you talked a lot about that, that's your AI event, but I'd love to hear the update if there any?

I think it's continuing to be the hot topic. The conversation has moved from consumption of tokens to optimizing token usage, not using tokens where not needed, using open weight where needed, building specialized models using open weight as the base, making a difference between using expensive close frontier models to not expensive close frontier models, capturing the learning and creating a learning loop and building an alpha around it. So I mean, all of these have become so much a hot topic of discussion, especially customers who are doing business operations. I mean software engineering is more mature now, business operations is not as mature.

So this is going to be a hard topic of discussion for the next 12 months, and it will actually lead to more and more work and more and more services for companies like us. And it also mean bundling pre-training and inference costs, along with our services, we will -- we now have arrangements with all 3 frontier model companies to bundle those services and bundle the inference and bundle the pretraining cost, which means we will have to build that craft and that craft is an important craft to build it because it will then mean that the input factor is not just going to be human effort. It's going to be human effort, platforms, software, frontier services, all bundled for an output, which is not effort-based but outcome-based.

Our final question today is coming from Bryan Bergin of TD Cowen.

I'm curious if you can share any kind of rough mix of the managed services business that already incorporates Gen AI-led efficiencies. And I'm just trying to understand the balance of the multiyear book that still needs to go through a cycle of renewals so that we can kind of better project a potential crossover point when you see a potential acceleration from AI activity that can more than offset that existing base compression and other factors.

Yes. I think it's -- as you know, our revenue has roughly 2 components, time and material and fixed price and fixed prices, both are now 50-50. So our view is that thin material is continually every time we renew it and that short-cycle business, typically 6 to 9 months, sometimes 12 to 15 months, but never more than 2 years or 2.5 years. So that's a short-cycle business. So that's continually embedding in itself, even on managed services basis, the benefit of AI into itself.

On the remaining 50%, which is fixed price book of business, typically, the contract lines are between 24 and 36 months on an average, of course, there could be some which are 5 to 7 years and some could be shorter. But typically, on an average, between 24 to 36 months. If we believe that we have started this journey of embedding AI into our solution more actively from beginning of last year, which is 2025, we are roughly halfway into it, and we have probably another half to go.

Okay. That's very helpful. My follow-up is on Project Leap. So just any further details, how much of the plan have you actioned thus far? Any kind of in-year savings from the program that you realized here in 2Q? And just anything important for us to consider as far as the pacing of kind of cost and savings yield as you go through 3Q and 4Q?

Sure, Grant. So we continue to execute the program. We have taken approximately $84 million of cost in quarter 2, of which $50 million, $55 million is from -- related with employee severance and remaining is from -- related with facilities and software and some of that. As we have -- as you know, we have baked in the savings from the program as part of our guidance range, and we believe we are executing well towards that goal. I think you should continue to see the rest of the year, evenly split from Project Leap execution between quarter 3 and quarter 4 as we move forward.

And we get full year benefit next year. And it also reshapes the cost of technology deployment in the market. To a large extent, this is about margins, but it's equally about growth, can we get more growth using a baseline where productivity is shared with the clients.

At this time, I'd like to turn the floor back over to management for closing comments.

Thank you so much for joining in today. We are very, very excited about our quarter 2 earnings, continue to be on the winners circle. We have confidence of staying at the winner circle for the rest of the year and create some nice tail velocity for the next year. And we are excited about the activation of all 3 swim lanes, productivity, doing old things in new ways, using AI. And as I talk, we are seeing accelerated momentum on new things using AI, which is primarily driving growth imperatives for enterprises. So thank you again for joining the call today.

Ladies and gentlemen, this concludes today's teleconference for Cognizant's Second Quarter 2026 Earnings Call. You may now disconnect or log off the webcast at this time, and enjoy the rest of your day.

I think we're ready to get started here. So I'm Surinder Thind, the technology and information services analyst here at Jefferies. And for our fireside chat today, we have Jatin Dalal, Chief Financial Officer of Cognizant. So welcome, Jatin.

Thank you very much for having me.

Fantastic. So for today, I've prepared a series of questions where I think the first half, we'll probably focus on AI, the AI debate. And then in the second half, I think we'll move into more of the questions around demand, near-term trends and kind of the shifts that are going on in the business.

Sure.

So I think where I'd like to start is just there is a lot of investor debate around AI and people are viewing this as potentially deflationary impact on IT services broadly. But my sense is there is an oversimplification here for a business such as yours, which you -- there's a broad range of capabilities and services and products and offerings that you have. And that provides different value pools, I would argue, for all the different clients that you have. So as you kind of think about this strategically, like how should investors disaggregate your portfolio and better understand where the AI opportunity is and maybe where the AI risk is in the business and how you kind of go about thinking about this?

Sure. So I'll start by saying that there is a -- there has been a sort of narrative around AI deflation, which is well known. And this is something that the industry has seen it for last, I would say, 18 months, where we have seen that the traditional IT services work that we do, we are able to do it more efficiently, more effectively with AI. And therefore, there is a deflation on the individual contract side, although there is very little deflation at an aggregate industry and company level because even with that deflation at individual project level, we are seeing that last year, we grew 6% plus.

This year, we have guided for similar growth rate for a range of outcomes possible. But we are -- the core part of that guidance is that organically, we will grow again 3% to 4% this year. So even with the deflation, there is very little contraction in the total revenue of the company or the industry.

But what is really exciting for Cognizant is that AI opportunity -- AI is opening up opportunities which are either to not available to industry in 2 ways. One, it is opening up new -- opening up old work, which you can perform in new ways. And second, you can do new work in new ways.

And let me talk about both. What do I mean by old work in new place? If you -- if all of us have heard about mainframe modernization, and that was the work that we used to do 15 years back, and now it's a tiny trail or small stream of revenue there. But suddenly, with AI, the mainframe modernization has become a proposition which is very real, which can get done in 24 to 36 months. It's no longer a 4- to 5-year journey. Certain upgrades of ERP is far faster to get deployed than what it would have been before.

So we just recently won a large opportunity on mainframe modernization, which is really not something that would have come as part of our pipeline any time before because of AI. So that's really old work, which is mainframe modernization, but now being done in a new way.

But let me speak about the new work that we do and the new opportunity that we are getting. And the easiest is to think about opportunity like Mythos. It didn't exist 4 months back. It is now here. But if you start with Mythos and then you discover, let's say, potential 200 critical vulnerabilities in your environment in application databasing network, you will have to go out there and fix it, and that opportunity was never existing when AI was not there.

The second is whole agent development life cycle. Traditionally, we have all lived around the ecosystem that was built around microprocessors. So we had microprocessor in the middle and then we had compute, we had storage, we had software, we had IT services, and that created the value for our large customers. So for 30 years, we have lived in a world which is -- which had only single direction, which was an ecosystem around microprocessor.

Now we are opening up the whole ecosystem around LLMs, which is LLM in the middle and there is a new compute of, let's say, NVIDIA and then there is -- there are a few LLM providers. There's a few agents provider and then there are branded agents, they are non-branded agents. And there is a whole training of agents and then eventual deployment of agent and then agent monitoring for potential drifts that can happen.

So if you envision that a large Fortune 500 company will work with 70% of the deterministic systems that it built over 30 years and 30% of probably stick systems that will be built around LLM. The whole runway for IT services company is around building and deploying those systems for large companies. So it's a new work to be delivered in new way. And therefore, we remain very, very optimistic and bullish for what we can do with AI as new opportunity.

So in summary, there is a pressure of deflation on traditional services. But if you see that is in our traditional space and that also we are able to overcome it by new work. And that's $1 trillion of TAM. But we believe this probabilistic system and deployment around it is another $5 trillion, $6 trillion of TAM that we are opening up with the advent of AI. So it's a great -- I would think it's a great opportunity for the industry and for Cognizant.

So from a messaging perspective, what I'm hearing is, yes, there's deflationary impact in the traditional business. But with everything that's coming down the pipeline, you've got a much bigger TAM to go after, and that's why you're able to kind of keep your growth rate positive at this point. Is that essentially the messaging that there's a lot more coming down the pipeline?

I would say there's a lot more coming down in the pipeline, but that a lot more coming down in the pipeline is being pulled forward on traditional IT services side. It is still not the new services revenue around LLM, which is yet to be seen in its full force by the industry. But when it comes, then you will have sort of a dual engine growth, one being driven by the traditional business and the one being driven by the new business.

Got it. And then as part of this kind of what I would call, you've talked about the shift away from the traditional services space towards more of a platform or an IP-enabled model. So at a strategic level, what does that evolution actually look like for the firm? Like how are you thinking about the forward business model? And what is this role that proprietary IP these reusable assets that you talk about that you guys are building. So is it more a software-like revenue stream that you're moving towards?

Yes. So traditionally, we have said that we have offerings that we deliver through skills, which is the pyramid of human intellect that delivers that services. Going forward, we believe that the world lives in a place where you will deliver the outcome in both the streams on classic IT systems and new AI systems, both you will cater to through 3 constituents. One is the skill. The second is influence or agents and the third is platforms.

And we believe that you need all 3 to deliver a most optimized outcome to a customer and any one is not going to be sufficient. And therefore, beginning of this year, we carved out platforms as a separate feeder of delivery to our customer. And within that, TriZetto is our flagship IP on healthcare side where 2/3 of the claims in U.S. are processed through TriZetto.

And that means it's a great IP because it has sort of thousands of type of claims on one side and another thousands of payers on the other side. And it has sort of a residual knowledge of processing the most complex to most simple medical claims in U.S.

Now where does it help us is that defined point A to point B journeys help us deliver far faster outcomes to our customers. And I'm using example of TriZetto, but that is while the carved-out platform is similar to that, we can -- we have IP in multiple other places, which can be leveraged to deliver an end-to-end outcome.

And one additional factor why we see TriZetto as a great IP is, it is suddenly with AI, the surface of monetization has multiplied significantly. Traditionally, I sold the IP on what I call as PMPM model as per member, per month model. And there are only so many members I can onboard on the IP every year in terms of growth. But we announced that we have now made TriZetto headless, meaning that any agent or any virtual -- sort of any virtual agent can come access that IP and get the work done and go out, which means now I have a monetization for the same IP with some incremental investments, but not significant. We have made it open to all the sort of agentic workflows, which needs to deliver the IP, which means I have additional monetization every time some agent comes and touches our IP. So that's an additional monetization reason. So that core platform is a big priority for Cognizant.

And the impact on margins, I think at the AI forum last week, the idea was, as you pursue this strategy, margin should be biased higher? Or how do we think about the impact on the business model?

These platforms by design have a higher investment and therefore, higher risk, and therefore, you anticipate a superior return in terms of gross margins. So therefore, yes, definitely, as the platform becomes part of a larger part of the portfolio, it should have a favorable impact on the operating -- gross margins and operating margins.

And then when we think about the next part of this platform or software strategy, like you've also highlighted other examples that you've had like a digital nurse or you've built like a wealth management agent. Now is the idea there that -- when I think about you highlighting this, the strategic significance of that, is there a lot more of that to come? Like how do we actually think about where you guys evolve to in terms of kind of this IT strategy versus a small supplement versus pushing really hard on that frontier?

Yes. So this is what I talked about, the new work in new ways. That is what we call Vector 2 and Vector 3 opportunities, which is coming because of AI, where we -- Vector 2 is making organization ready for AI and Vector 3 is actually deploying agentic workflows for our customers. So either having a wealth manager agent or an agentic nurse is a classic agentic deployments with agent development life cycle and monitoring of that deployment over a period of time.

And that's absolutely a new opportunity that we never had before. And it is, therefore, a big focus area for us. It's a tiny revenue stream today, but if the growth that I spoke about, which is from $1 trillion to $6 trillion, the large math of that growth is the new ops that we never did before, which we can do now with agentic workforce is through opportunities like this.

That's helpful. And then just in the very recent, there's been a lot of talk about tokens, token economics, the spend and all the challenges that clients are having. But as you yourself consume more AI to deliver your services, more compute, can you help us understand the implications for your revenue model and how that -- how you work with clients on that?

I mean, is this an idea where you will just -- the client buys the tokens and you build on top of that? Is this a situation where maybe you guys buy the token and they cost plus it? Or do you just go down the whole path of fixed price where the client just -- they don't care. They want a certain service for a certain price. How do you think about those revenue models and where you end up in this?

Yes. So I think we are at a point where our customers have just begun consuming tokens materially. Of course, there are cases where people have overconsumed tokens. I'm sure many of you know about this. And therefore, there is a question now as to the throughput or the value created by tokens or inference, right? We work with our customer on all 3 models that you mentioned, meaning I create the Tier 1, which is agentic development life cycle and I actually create an application or agentic deployment for customer, but I only charge for human effort that was there to deploy it. But that's a smaller component.

What we are increasingly seeing is customer is asking us to blend the inference as part of offering. But still, it's very transparent with customers how much inference is getting consumed. They're willing to let us make margin on that, but they would like to have visibility on that.

Eventually, we see a fixed price project that you type of deal or outcome-based type of deal where customers says, I don't care how much of platform you're using, how much of inference you are using, how much of skills you are using, you tell me if you can deliver this outcome at this price to me, then I will buy it. So that could be -- that's a sort of, I think, somewhere in future, that model will start getting traction. Right now, we are at a point where we are beginning to embed inference in the rate cards, in the fixed price project. It is visible to customer today. It's not very tight fixed price. But we -- the evolution has already started.

Got it. And I've also heard you mention in some very recent conversations that the fact that clients are questioning their token usage, their budgets that is perhaps a leading indicator of change or demand. Can you talk a little bit about that and how where you might help clients cross that bridge?

Yes, absolutely. I think there is a broad realization that there is a cost associated with token and therefore, there has to be -- like any other cost, there has to be some amount of prudence on how you token cost gets consumed. Where Cognizant comes in play is just the way we have traditionally built a pyramid of human intellect and skills, we build a pyramid of LLMs and SLMs where not every query needs to go to the most expensive token. You could essentially build a small language model for your accounts payable process in your office, which caters to 70% of the users' queries or agents' queries and agent can freely work with that.

For the 30% of the queries you need to filter out, will go to any of these models. That is fine. You don't need to send every possible query to the -- to an LLM. In a very industry-specific domain, you don't have to work with a large language model. You can work with a narrow language model. So you could really optimize your intellect usage depending upon the query that comes in. And that's where we add value. And that's where we become relevant when somebody feels that they are overspending on tokens versus the value that they're generating from it.

Got it. And you mentioned something interesting here, which is this idea of introducing a lot of -- it sounds like proprietary models. Can you talk about that versus this idea that maybe there's 1 or 2 big winners? And how do you see that evolving and maybe the investment that you're making in building your own industry-specific or workflow-specific models?

Yes. I think there is a certain amount of prominence that a few models, large language models have got because of their universal applicability, but there is even today, availability of narrow language models or what we built for our small language models for our customers for the limited domain that we do. Typically, it's their IP, because it's very relevant for that customer. But it makes the whole ROI question, very favorable for customer versus the question mark that sometimes comes when for simple queries, you are using certain things.

Also, you must remember that there is always a most optimal way of doing a few things, meaning if you want to add up 15 numbers, you're going to go to your Excel or your calculator. You are not going to go to an agent and ask the agent to compute. And so it's a very simple example. You could extrapolate that to everything that you do in an enterprise where you don't have to go to agentic answer for everything. You will have 70% of work being done by today's systems and 30% by, let's say, agentic outcomes. And that's where we add value, where we marry the 2. You are not solely relying on one type of delivery to get to an optimal answer. So I think the question of token cost and token usage is relevant. And I think that's where companies like Cognizant can add value.

Got it. And so maybe just putting together the last few questions in kind of one big wrapper here. When I -- I think the idea is -- I'm trying to get to is what does Cognizant maybe look like 3 or 5 years from now, right? I mean, if enterprises start to move meaningfully from kind of what we have is deterministic systems with human wrappers, right? That's the existing workflows towards more of these autonomous agents that operate across the enterprise. How does this -- how do you look -- where are you in that vision 3 or 4 years from now?

Our vision is -- comes from what customer requires. And we believe a large Fortune 500 customer would be using maybe 70% deterministic systems that we use today and maybe 30% would be probabilistic system built around LLM. And we would have a meaningful role to play on both sides. We have sales capability for probabilistic system. We have sales capability for traditional deterministic system. We will have delivery, as I mentioned, with a combination of human skills, influence or agents and platforms. We will have a model of Cognizant behind it, which would be far more in our assessment, flatter and wider, where multiple middle layers of the organization may not -- may be merged in a way that it creates more value for our customer. That's what we see as a model of the future for Cognizant.

Got it. And then I think the final thing here, just maybe I want to talk about the pace of change, right? And I think the idea here is how quickly it is. If we go back to just Anthropic blog post last week about how things like the software engineering capabilities of models are advancing maybe faster than people are going to be able to keep up the length of the tasks, everything that's going on.

How do you actually plan and commit to your investments when the world around you is changing so fast? I mean, let's say you were working on something 6 months ago and all of a sudden, there's new capabilities introduced. How do you work through that and plan around all of that?

Yes. So we have -- so I'll answer it in 2 parts. One is how we keep up with that. I think, we have probably the best in the industry lab for AI and probably some top minds, researchers work for us. Our Chief AI Officer is probably one of the most respected names in the Valley. So that's how we keep up with the pace of what's going on.

But the more relevant question for all of us in the room is, there is a technology potential, which is, let's say, 100. And every day, it's growing. So in 10 days, it will be 110. I mean, I think there is a new model, which has been put up by Anthropic this morning. But if you see the value that enterprises have delivered, there is nowhere close to 100. I mean, you all may have a different number, somebody will say 10, somebody is at 15, somebody is at 20, but it's nowhere close to 100.

So I think there's a big value gap that needs to get bridged. Because eventually, every technology is only as useful as the eventual value delivered by the end user. And I think there is that gap that is visible today. And organizations and companies like Cognizant who are primary advocates for our customers to switch the ecosystem well and deliver that value because if technology continues to progress, that's great. But I think today, the bigger and burning question is, how do you bridge that gap between technology potential and value to the enterprise.

And so I think that will lead us to the next section here, which just the broader demand, demand environment. So when you talk to clients today, right, I think there's this concern about what I would call mounting like legacy complexity. There's a lot of technical debt. Things are changing fast. How do you characterize that conversation that you're having with clients today, right? It seems like there's a lot of demand, but at the same time, there's a lot of concern from a client perspective. So can you help us understand what's going on?

Yes. I think there is -- and that brings us to a little bit on short term. I think we definitely see that on one hand, there is a lot to do for our customers. But at the same time, the world around us is changing so fast that is creating a certain amount of indecision or delayed decision because nobody wants to jump into something which can be later called a technology of yesterday. So there is that pause or delay that we have definitely observed. And -- so while currently, the traditional work is being done far more efficiently with use of AI, the new work related with AI will take its course as it continues to build momentum around itself.

And then I guess at this point, like what would it take to get clients to engage maybe a bit more aggressively or invest a bit more aggressively? Like discretionary spend has been weak, broadly speaking, for a number of years now. Can we talk a little bit about what's going on there? And do we need to see maybe some stabilization in the technology to kind of get clients over that cliff? Or like are we just kind of stuck here for a little bit?

So I would think there are multiple data points which point towards saying that we are taking maybe not fast enough steps, but we are taking firm steps towards AI adoption. If you see some of the revenues of large LLM players, they are growing very rapidly, which means that large enterprises are consuming that. And when large enterprises are consuming that, they are all smart buyers. At some point, they would start investing effort on how do we get ROI on that. So that will come.

So I certainly don't see -- I don't think that you need to do something different to land there. I think, we are taking a slower step, but firmer step towards AI adopted world, whether it is -- and a result of that discretionary business for IT services industry is 2 quarters away, 3 quarters away, 4 quarters away, difficult to call, but I definitely see ourselves working towards that.

And then if I could maybe ask for a bit more color on that. You've conceptualized it in this concept of Vector 1, Vector 2 and Vector 3, right, where Vector 1 is productivity led, Vector 2 is more about infrastructure and the build-out of all the capabilities so that you can use AI. And Vector 3 is where you effectively redesign workflows or you go AI native. When you look at the mix of demand right now, where are we there? Like are we seeing enough signs that we're moving between the vectors at this point or...

Yes, I think it's a great question. I think we definitely see -- still a large component of our pipeline is Vector 1, which is productivity-led IT work. But we are beginning to see a meaningful pipeline for Vector 2, which is making enterprises ready, especially on data side, which is Vector 2 work. We are seeing first implementations of agentic workforce, which we spoke about in our AI Forum on Friday, where customers came and spoke about it. So that's also beginning to pick. But still that is relatively small and some way to go. But Vector 2, I certainly -- I'm beginning to see a good traction on Vector 2.

And does it all have to move in sequence, meaning you got to do Vector 1 first, be then, I would say, dissatisfied with the level of productivity that you're getting. So you're like, I got to rebuild my infrastructure, then you go and rebuild it, get your data cleaned up. And then you finally kind of move to Vector 3 or like, who knows Vector 3 today. So...

Yes. So again, a great question. I don't think you need to go sequentially. It depends on the architecture of your enterprise, whether you need to go sequentially. AI is very powerful. Even with a slightly suboptimal data structure, I think agentic deployment will work well. But it may be a little bit more expensive because now inference has to work through navigation of complex data structure. So it is not sequential in some form.

But if you want to optimize each aspect of your agentic development, then you would want to optimize your data structure and then go for the inference cost rather than inference do extra work on data side too.

Also, there could be companies which have invested 2 years back on a great data structure. They don't have to go to Vector 2. The companies which are using agentic workloads effectively are simpler organizations, what I'll call single geography, a large line of 1 or 2 businesses. They are effectively deploying agentic workforce as of today. It is not a conversation of future. They have replaced human efforts with 40%, 50% of virtual effort, and they continue to do so. If your business is a little more complex, you have 300 products operating in 50 countries, that is taking a little more time before they deploy Vector 3.

Got it. And then for the -- what I would call the native company or the ones that have kind of tried to make the Vector 3 type of transformations, how are they measuring the benefits? Like are they seeing what they're supposed to be? Because there's this mixed narrative of it's really hard to right now realize the benefits of AI. Are you beginning to see signs where we can start to measure some of those benefits and that ultimately becomes one of the situations where when we look about past cycles, it's the success that one client, they build a competitive advantage. And then all of a sudden, the competitor says, wait a minute, if they did it, I also need to do it. And so you kind of see this acceleration.

Yes. I think where your agentic deployment, the answer is very comprehensive and very clear. It's a question -- I mean, it's a night and day difference. I mean, one of the customers who spoke in our AI Forum spoke about saying he had a few hundred agents who are performing a service. He has moved to agentic model. The number has come down sharply by -- I mean, he's left with maybe 10s or 20s of those and effectively being replaced by virtual agents.

The turnaround time has come down from a few days to a few minutes of the process that they were doing. And what he takes -- talks very proudly is that of his few hundred agents, the virtual agents he has trained are the -- he picked top 5 agents from those few hundred, and he trained the virtual agents with those best performing agents. So now he has almost entire throughput going through virtual agents who are trained from his best agents. So his quality of outcome has increased significantly. Number of days of turnaround has become a few minutes. So it's a comprehensively superior outcome.

Got it. And then maybe shifting a little bit to -- I don't want to get too near-term focused, but there's a lot going on in the current environment when we think about it from a geopolitical perspective of where we started the year, there's a lot of excitement. Maybe can you talk about just the evolution of client behavior and maybe like the last 90 days, just kind of where we sit in this macro geopolitical uncertainty and how -- like at least from your perspective, like -- what are the changes that you're seeing from your clients in terms of their decision-making cycles, maybe prioritization around projects or just maybe the visibility that you have to your client spend relative to historical norms and standards at this point?

Yes. So I mean, we spoke about in our quarter 1 earnings call that there is a near-term macro uncertainty, which we see in customer behavior. It is driven by geopolitical drivers. It is driven by a little bit of rapid change in technology that we spoke about earlier. And things have not changed since then, meaning we still don't see any acceleration on demand side from that situation. We have guided for this quarter, keeping in mind that environment, and that's what we think is playing out. We had also budgeted for, within our guidance, a month revenue from our recent acquisition, Astreya, but that is yet to close. And once it closes, we'll make an announcement around it.

Got it. And then when we think about just where -- are you seeing any geographic differences? Because there's also a concern about how the different economies are evolving, let's say, Europe versus the U.S. or even Asia Pac at this point.

In our largest geographies, which is both U.S. and Europe, we see a growth momentum, which is quite uniform. Especially in Europe, we -- when Ravi took over as the CEO of the company, U.S. was among the first to start showing the traction and the growth. From -- Europe took a little while, but we have had some marquee events in Europe in the last few months, and we remain quite optimistic for 2026 for Europe, too.

And then I guess the final kind of component here within the current environment, your BFSI segment doing really well. They're spending, they're ahead of the curve. Can you maybe talk about what the -- what's going on there versus is it just a macro issue? Or are they just kind of -- are they just leading the space in terms of the investment here versus maybe what's going on in like retail or healthcare at this point?

So BFSI is always a leader in appreciating sort of potential of new technology and its deployment. So we certainly see BFSI leading the whole investment on AI and therefore, the discretionary spend related with Vector 2 that is quite visible to us in BFSI space.

I would say following that is health. Manufacturing is doing decently okay, but probably most impacted from the geopolitical uncertainties. And finally is communication and media space, which also is not in its prime in terms of growth. There are company-specific sort of priorities that our customers are tackling.

And broadly speaking, you've had a lot of success in winning a lot of large deals. That said, on the discretionary side, there's been a bit more softness on discretionary spend yourselves industry-wide as well. Can you talk a little bit about that dynamic? And one of the narratives out there, there's concern that as these tools advance, the AI models advance, clients are trying to do more themselves. And so that's kind of keeping discretionary. How do you think about what might be going on there?

Yes. I mean -- yes, I mean, I have not seen clients doing more of new work with them. It does tend to happen because any new technology client wants to have a comfort and feel of operating it before it gets outsourced. But I don't think that's been a driver this time around. Even we have had very large successes with setting up GCCs for our customers. So even customer GCCs, we are helping them run in India or elsewhere. I think it's just -- outside BFSI, it's more macro and the fast pace of AI tech evolution that is probably pausing decision-making a little bit.

Got it. And is there any indicators that you're looking for, meaning more positive client conversations or commits? Or how do we think about somebody that maybe sits from the outside trying to understand when we kind of get over this hump, right? I'm not talking about predicting next quarter that we have, but just kind of how we think about it broadly.

I think it's -- essentially, how it starts is the texture of the business changes from many large deals to multiple small deals every time there is a discretionary rhythm change. And I have a feeling that when it changes, it will be very clearly visible. We speak also about our ACV growth numbers every quarter. And when you see ACV growth, maybe a couple of quarters inching up, I think that would be a great indicator that finally, we are seeing the discretionary back.

Got it. And then I want to -- there was an interesting comment that Ravi had made about just kind of right now, you have peer-leading growth, but then you talked about getting to like breakaway growth. Can we talk a little bit about that, the aspiration there and the genesis of that? Like when I think about it strategically, financially. Like why lay that goal out there right now? You're already peer-leading and now you've kind of set another aspirational goal about getting to breakaway growth.

I think the command for breakaway growth is in light of the opportunities that AI builder or Vector 3 provides. And I think if we get to that zone, I think a higher growth is definitely possible because we are going after a much larger TAM than what was visible before. So that's a statement of aspiration that as we move from a traditional $1 trillion TAM to a much larger TAM, you would accelerate your growth rate significantly.

Got it. And then just kind of wrapping up here, maybe something on just kind of capital allocation and M&A. Can you talk about just your capital allocation strategy at this point? And how M&A fits into the broader strategy at this point? And then I've kind of got a couple of follow-ons about the strategic use of it.

Sure. So our capital allocation philosophy is 50% of the free cash flows for M&A, 50% for returning to shareholders. Within that 50%, half is for dividend and half is for buyback. Now last year, we did not have a large M&A. We were consolidating Belcan that we had acquired a year before. So we had a large $1.4 billion of the 2.5%. So roughly 60% was our buyback.

This year, we started with slightly higher allocation for buyback because the share price was lower. But in month of May, when we realized the share price has dropped further, we went with another $1 billion of -- so collectively, $2 billion of buyback is what we have planned for 2026. And that, as you can see from -- it's nearly 80% of the $2.5 billion of free cash flow that we generate. But the way I see this buyback of this year is more a pull forward of our maybe future years, because the time was so opportune that we had to act and we acted. But over a period of time, we would like to retain this balance of 50% of use for M&A or strategic use and 50% for returning cash to the shareholders.

And then in the current environment, I can understand the share repurchase component of it leaning into it. But what about this idea, given the amount of change, leaning a lot more heavily into the M&A, right, to find new ideas to kind of diversify against all of the different possible outcomes that there are? Like are you still all in on M&A.

Yes. I mean, one could go all in on M&A. And I agree with you, these are the times where you should lean into M&A for going out of new opportunity. Both acquisitions we have done since the beginning of this year, 3Cloud and Astreya, really play in that whole AI super cycle. 3Cloud really works on the cloud deployment for AI workloads and Astreya is right in making data center ready technologically, and we all know the kind of investments which are going in data center.

So clearly, we are picking up businesses which help us gravitate faster and more substantively towards the AI opportunity. And we'll continue to look at that. I mean, a good thing about the current market is that, good companies or good opportunities are available at reasonable prices. And we'll go -- we'll continue to invest that. But you also balance where you are in your share price and therefore, repurchase made imminent sense at that price. So we went all out. But that doesn't -- even as we made it, we were quite categorical to make a statement that, that doesn't restrict our M&A option. We have plenty of leverage opportunities or leverage bandwidth on the balance sheet if we found something big that we can't manage with our cash flows.

And then final question. I noticed we're down to the last minute or 2 here. Anything you want to lean into the portfolio from either an investment perspective or an M&A perspective where you'd like to add to something maybe? Like I feel like you've made some important product engineering acquisitions. Anything that sticks out to you or?

No, I think platforms is one. Operations is another because the whole hypothesis around targeting tech plus operations could be another area of investment or interest to us.

And one last point I would make on investment is that we also announced Cognizant Innovation Network, which is really the investment in early-stage companies, which bring that additional innovation edge to Cognizant. An idea is that we invest in that IP and we bring that IP to our Fortune 200 customers. So there's an additional investment. The dollars are not that big, but it just provides a very different ledge of innovation to go to customers with. So yes, I think, these are the few ideas to double down on.

Okay. Well, fantastic. I covered a lot of topics here. So I really appreciate the time.

Thank you very much. Thank you.

All right. All right. Thank you all. Thank you all. Good morning, everybody here in the room in New York. Before I get started, Go Knicks. Anybody? Yes. All right. A special hello to everybody on the webcast as well. Thank you all for joining us, and welcome to Cognizant's 2026 AI Forum. My name is Tyler Scott. I'm the Senior Vice President of Investor Relations, and it is a pleasure to welcome you here today.

We're excited to be joined in person by many of our industry analysts, advisers, clients, partners and of course, some of our shareholders and sell-side analysts. So thank you so much for being here. Today's AI forum addresses a defining question facing businesses right now. How do you close the gap between the AI promise and the potential and the enterprise-wide impact? Last year, we put our AI builder strategy in motion and bridging this gap is exactly where Cognizant operates.

This morning's keynote session is going to be webcast live with our CEO, Ravi Kumar; and our CFO, Jatin Dalal, and a replay will be available on our website after this call. For those of you joining in person, we encourage you to take advantage of the networking, demos and engagement opportunities in our sessions later today. And now, of course, one of the most important updates, while we're not going to be providing financial updates, some of our comments on today's webcast and in our meetings today will be forward-looking.

These statements are subject to the risks and uncertainties as described on the slide behind me and of course, our other filings with the SEC. With that, let me welcome our CEO, Ravi Kumar, to the stage.

Thank you, everyone, for making it today. The weather in New York has got better in the last few days. So very excited to host you all. We have a very diverse audience, sell-side analysts, industry analysts, some of our partners, investors. And most importantly, while we speak about our strategy, in fact, when we got to an Investor Day last year, we spoke about expansive margins being on the winner's circle and our EPS growth being greater than our revenue growth.

So that mission and that aspiration continues to be with us. Most importantly, today, we have our clients who are going to talk about some of our early success, learnings, early wins and the partnership with Cognizant. LPL Financial, Citizens Bank, Signature Performance, Elevance Health, U.S. Bank, Lineage Logistics, Sysco Foods, TD Bank. So we have 8 clients talking about early learnings, early wins with AI.

Every company and every industry is set for a purpose. It's set to address a white space, a problem -- a purposeful problem. And companies are built on a mission around it. For 2 decades, our industry, Cognizant, built a mission around expertise to navigate enterprise transformation with the power of technology. Classical software manifested into systems, and we globalized and scaled companies by applying those systems to business flows.

And we built a craft. The craft was about building these systems to scale enterprises to efficiently and effectively run and we could do that better than our clients because of the repeatability and the rinse and repeat of what we do for multiple clients. And we embed that into software engineering for systems and products, and we kind of got into the adjacencies of managing the infrastructure, managing the process, managing over a period of time, infrastructure went virtual.

So we managed to transition ourselves through these tech disruptions and kept that unique craft of helping clients go through that transformation. We transitioned ourselves from being a system builder to a systems integrator over the last 25 to 30 years. Here we are now with this extraordinary AI technology. It's a platform shift. Internet scale data, neural networks and an extraordinary compute all coming together. And it kind of blurs the line between systems and people to run business flows.

And because it blurs the line, we have a much bigger sprawl beyond systems to operations. Look at the economic data. This is a tailwind of what we do to the AI opportunity. $20 trillion of economic output by 2030, of which $6 trillion is going to be productivity, $14-plus trillion. So 1/3, 2/3, $14 trillion is about new products, new services powered by AI. A lot of what you see today is productivity led. And this is going to create some kind of a flywheel. There's $1 trillion of spend on the AI infrastructure.

$6 trillion to $7 trillion is going to be spent in the next 3 to 5 years. And effectively, that output is only going to come if the capability of AI, which is right out there and it's going higher and higher, it's only going to get better and better and the production value to enterprises is way below. And that is the gap we are trying to address. $1 trillion already invested, another $6 trillion to $7 trillion to be invested by 2030. The capability is going up.

The production value is right out here. In some ways, that's the gap we are going to address as a company. The costs are ballooning with very little productivity. And we're already seeing enterprises talking about it. The challenge. I mean MIT spoke about it. McKinsey spoke about it, Bain spoke about it. Uber put it on the list on their earnings. Microsoft spoke about it. I saw a report of Walmart actually speaking about how this is becoming a challenge.

Sam Altman last week spoke about how he's already seeing budget issues related to AI in enterprises. So there is a lot of reasons why the production value is way below the capability. If Chapter 1 was about broad open-based experimentation, Chapter 2 would be about specifics, realism, cost and bridging that gap. And here, we are building a new category of a company, which is going to evolve from the past, some craft from the past, some new craft, we would now build this new category of a company with a new purpose to bridge that gap.

And that is what I think the opportunity in front of us is. All right. So we did this study called New Work, New World in 2023, and then we did a revision of it in 2026, 18,000 tasks, 1,000 occupations from the O*NET database. And we actually projected that 10% exposure to AI, 93% of the jobs would be 10% exposure to AI. 25% exposure to AI is 69% of the jobs and 50% exposure to AI, 50% of the jobs. This 30% of the jobs and this specific exercise which we did, we thought all of that is going to happen in 2030.

Here we are in 2026. It's already there because the velocity at which this is going is significantly higher. So we did exposure scores, friction scores and velocity. And what is really happening is systems plus people, which created flows and our universe for systems is now becoming systems plus people plus digital labor or Agentic work. And the universe is completely now spread across this entire spectrum.

And the lines between systems and people are getting blurred. And I'm going to give a little bit of a back of the envelope math on this. And that's where I got this $4.5 trillion too. The total global economy is roughly around $120 trillion. Global 2,000 companies have $52 trillion of revenue. $20 trillion is operations-related spend. $1 trillion is spent on systems by the global enterprises, G2K. They spent $1 trillion on IT services, system integration work. They spent $350 billion to $400 billion on software, enterprise software. And now the universe we are talking about is $1 trillion of IT services from systems. We'll talk about it. It has a new flow. And we have 1/3 of the $15 trillion to $20 trillion, which is invested on operations of companies by the Global 2000, which is $20 trillion, $4.5 trillion is the labor exposed to Agentification.

So the universe we are talking about is $5 trillion to $6 trillion. So it's a 6x opportunity to the opportunity we actually chased for the last 25 years. That $6 trillion of -- an opportunity is going to be labor and business operations of companies, Agentification and reviving the systems to do that. And when we do so, you're going to create a flywheel because as the labor gets Agentified, the human endeavor is going to look for new purposeful work. And that flywheel will create consumer spend and the $20 trillion of economic output will create more wages and more jobs.

And we think those jobs of the future are going to be significantly higher than the jobs of the past. In fact, there's a World Economic Forum report, which talks about how 125 million jobs are going to be created by AI by 2030, while 70-odd million jobs are going to be knocked away from the past. So what are we trying to do? We are trying to find new value pools in this new universe of $6 trillion. And that unlock of the new value pools is what we're going to look for as the opportunity to bridge that gap and the opportunity to add value to actually bridge that gap from capability to production value for enterprises.

The pace of change has been absolutely high clock speed. There's been a sense of FOMO, fear mongering, and that has led to token consumption without linkages to ROI and without linkages to outcomes. So one of the reasons for this gap between capability and production value is also because there has been relentless token consumption without linkage to outcomes. Let's take the frontier model companies, roughly $100 billion of annual revenue as of today. They're probably going to do $1 trillion by 2030, $500 billion to $1 trillion. If the scale-up of the infrastructure continues the way it is continuing, now let's take a more optimistic scenario, $1 trillion. That revenue of $1 trillion because of the effectiveness of how the token consumption should be, a part of it is actually going to be routed through system integrators or AI builders, as I call it, like Cognizant.

Why? Because you need to create more efficient, more effective. Remember, this is a contextual science, more predictable and better economics for the token consumption. So Cognizant is actually building a token and an agent harness, I call it, with model operability, interoperability, the ability to capture the patterns of work, the ability to wire the digital traces and actually create a compounding factor for the digital labor to be routed through us.

We do this on a rinse and repeat for hundreds of clients. So shouldn't we be doing better than our clients because we do this much better because we do it hundreds and thousands of times better than more than our clients. So our ability to capture those digital traces. Remember, when we did that with digital human labor, you came to us and you said, "Oh, you did this at 100 clients, why don't you do it for us? We are using the same philosophy to get digital labor to be a part of our integral ecosystem.

Token spend is an architectural problem as well. Context quality. You're going to hear about context a lot today. Models don't know which context matters. They process whatever you give them. So you have the ability to filter the context with quality, model routing, not every step in an Agentic flow requires frontier reasoning. Continual learning.

Enterprise AI systems shouldn't solve the same class of problems from scratch every time you actually get them, harness design, all of that put together, especially because operations work is multistep, you can actually capture these digital traces, integrate it with human effort and create a new craft, which will make token economics an important part of your value chain for companies like Cognizant to not just help you externally as a partner on human labor, but also on digital labor and the integration of human and digital labor.

So here is what we are looking at as the opportunity. It's a $6 trillion market, not a $1 trillion market, and we'll come to the $1 trillion market on systems, which we have been working on for the last 25, 30 years. Software engineering integrated with Agentic into business operations, business flow reinvention, business flow reimagination, enterprise AI guardrails, context engineering. We spoke about context engineering since 2024.

It's the new craft. Writing code was a craft we built. Assembling context is going to be a new craft we are actually developing, and we have a partner with us who we work with, you're going to see in the later part of my presentation. Token harness, agent harness, workforce -- human workforce and digital workforce integration, outcome-based services. That is going to be the mix in the future of what companies like Cognizant, we call them AI builders will do in the future.

And that's the reason why we think we can bridge that gap between production value to enterprises and AI capability, which is continuing to go up and up. The bridge to production value gap is a 3-vector strategy. I've spoken about it since almost 2024. Vector 1 is software engineering, rebuilding -- I mean, classical software was done in a way we are going to now -- and in fact, as we talk, we have made all our software engineering autonomous asynchronous Agentic with humans in the middle of the chain as software engineering.

In the last 2 years, we used that opportunity to create [ more for less ]. And the paradox of living for [ more for less ] gave us the opportunity to push the backlog of our clients at a lower cost, higher velocity. It also gave us opportunities to consolidate. In fact, every quarter, we did a large number of $100 million deals based on consolidation, but it also created throughput for increased software spend. Every time this has happened in the last 50 years, when technology has got cheaper and technology has got easier, more technology got consumed.

And for a moment, even if the elasticity doesn't exist, you still know that there is a $5 trillion of new spend in operations of companies, which is getting unlocked because systems and people are now getting blurred and systems and people are now -- systems and people and digital labor. So we see the Vector 1 opportunity by itself on its own standing for a higher velocity and a higher elasticity of spend.

And I'm going to talk about some of the value pools. Vector 2 is all about industrializing AI, which is integrating AI into enterprise landscapes. Unlike in the past, where we took the old technology and junked it, this is a technology which is going to layer on top of the old technology. You could argue whether the older technology will become a system of record or it will be less consequential, but it is not going to be taken away.

It's going to sit on -- the new technology is going to sit on top of the old technology and integrating it, securing it, building the data funnels and integrating it with the system of record and creating the interplay between the classical software and the Agentic software is Vector 2 and Vector 3 is the sprawl into operations of companies, which we believe is 20% to 40% of spend of enterprises versus the tech spend of enterprises in the past, which was only 0% to 10%.

So this is how the production value and the gap is going to be fixed as we go forward. In a simplistic way, I see it in 2 swim lanes: Old things done in new ways, and new things done in new ways. Old things done in new ways is classical software with higher productivity, autonomous synchronous way. New things in new ways is this Agentic sprawl on business operations and building new products and new services.

I mean we have not got to it yet, right? The $20 trillion of economic output I spoke about by 2030, 1/3 of it is productivity. That is what we are actually now seeing in AI. As you go forward, you're going to see new products, new services. And when that happens, you're going to see a very different kind of growth imperatives for companies attached to AI and consumer-led -- AI product-led consumer consumption, which is going to come into picture.

So here are the value pools. It's a little busy slide. The first block is value pools of old things in new ways. We spoke about autonomous software engineering. In addition to that, this is something I'm seeing today. This is not a futuristic thing today. And this list will evolve as we go forward because this is running at such a high pace. If you -- if I do this presentation in 3 months, I'm going to show you a newer list.

Now let me quickly run it through. And today, for the rest of the day, you're going to see these value pools highlighted by my colleagues in more detail. First, secure AI services. It's a Y2K moment for me. Imagine doing vulnerability discovery using Mythos or GPT 5.5 and you do that vulnerability discovery at machine speed. That's not the opportunity.

The opportunity is to patch, refactor, remediate these applications. And that's why I call it the Y2K moment. When the Y2K happened, we opened the lid to fix the dates, and then we found a whole bunch of things, architectural deficiencies, brittle logic, and we tried to fix it. That is what we're going to see. It's a 10x opportunity. And people -- clients will be paranoid because that vulnerability is exposed. So we have partnered with CrowdStrike, Palo Alto and Zscaler with our security services, and we've got the entire company's service lines to rally behind it.

Modernization, mainframe modernization, I've already started winning deals. We have a huge pipeline on modernization. 800 billion lines of code of mainframes, lines of code have written, and they've been in enterprise landscapes as some of it as legacy debt for more than 50 years. It used to cost us $10 per every line to refactor. It now costs $1.5 to refactor.

SAP HANA migrations, SAP changed the dates every year because the ROI to do that was lacking by enterprises. Here we are now, we could do it faster, quicker, cheaper. SaaS reimagination. There's going to be some SaaS reimagination. Either you're going to apply AI on top of it or you're going to take some of that deterministic logic and push it back into SaaS. Autonomous infrastructure using AI. In fact, it's actually one of my blockbuster offerings. Autonomous -- running high-touch autonomous infrastructure even for the old stuff, leave alone building AI infrastructure for the new stuff is a -- it's actually a blockbuster offering.

In fact, as a company, we are already doing double-digit growth in that space. Cognitive Agentic commerce based on desire and intent. SaaS-enabled AI. Now AI has this unique thing to sit on top of SaaS or go on the side and you actually do Agentic work and open the SaaS layer and make it headless so that AI can access the same set of rules. By the way, we did that for TriZetto last week. We made TriZetto headless. $0.5 trillion of claims go through our TriZetto platforms. The same set of clients have $4 trillion of administrative work sitting outside of TriZetto.

So we made that headless so that the Agentic you do, you can do it on top of TriZetto, you will be super happy. If you want to do it outside of it, you can access the same guardrails, the same security, the same trust layers, which we built in TriZetto. That's going to happen on all SaaS platforms. I spoke about context engineering. This is a contextual science. The new code is actually to assemble context. We have spoken about it since 2024, and it's now getting to reality.

You are going to hear some of our clients talk about it. Sysco Foods actually implemented this. What is context engineering? We took all the deterministic parts and wrote classical software, codified it. We took all the probabilistic parts, and we put it in a human endeavor. Now we are saying this is a technology, if you ground it well into the hustle of the company into the heterogeneity of the company, you can actually create a science around it. That's the new code we are writing.

Model engineering, LLM versus small language models. Sometimes small language models are more superior. In fact, my data analytics teams are working on small language models for some of our clients. Domain-specific models and digital twins replicate what a human does and then plug things out of the digital twin and create an equivalent of view and then integrate it.

Physical AI, it's not taken off, but it's going to take off. It's a much, much bigger opportunity because the leapfrog for digital enhancement into physical, robotics, autonomous systems, mobility is going to happen. We have now invested heavily to seize that opportunity as it comes. AI infrastructure. Earlier this week, we actually put a press release on our AI factory. And this year, we are Dell and NVIDIA's the AI velocity partner for building AI factory.

Trust and regulatory guardrails. Agent development cycles are going to be very different. If you're building agent development cycles for new products, they're very different to the software development cycles. Agent development cycles, you scope work, you scope for outcomes, you design for behavior, you do iterative learning loops to ground the agents and then you supervise agents. You don't keep the lights on. So ADLC is a much bigger service. AI data training services. Cognizant's BPO arm, which is called the Intuitive Operations, has been doing machine learning data training for the last 10 years on the big tech companies.

We are taking the capacity and repurposed it for AI training, model training. And that's a huge service we are excited about. So what is the reforge of first principles we need to do? To be that company which bridges that gap between production value and AI capability. The reforge of the first principles are we go from a system integrator to an AI builder, and I'll come to it in a minute what an AI builder means.

It's a bespoke opportunity because this is not the deterministic piece. This is the probabilistic piece. So it's a bespoke opportunity, much, much bigger and the capability is very different. Pyramid to interdisciplinary talent. We thrived on a pyramid of talent. You're going to hear from Jatin on how that is going to spread. You're going to hear from Kathy, our CHRO on how that mix is changing.

Interdisciplinary talent at the intersection of domain and technology at the intersection of operations and technology. That's what we are trying to put and then embed digital labor into it. platform-led Agentic delivery. Earlier this year, we launched the platforms unit. Our thesis is our clients are going to consume this not as services, not as software, but as platforms, the pioneering more forward-looking clients.

I'm going to consume this as give me the outcomes through a software plus people endeavor. So we have an opportunity to be a platforms plus people company, and we used -- we have been a platforms company in health care for many years. And we need to have the courage to own business outcomes because no longer are we building systems, we are building -- we have blurred the line between systems and people, systems and operations. So this makes our business more durable, sticky, more value capture, a different risk profile and a different return profile.

What does it take to be an AI builder? What are the blocks, building blocks? Full stack bespoke capabilities, integrating human and digital labor into software engineering and business operations, flatter organization structure. This is the transformation we have to go through. We think the people in the middle are player coaches. The nodes are going to be more real time. So the people who actually were moving information up and down, coordination, measures, orchestrators, we shrink that.

We create more flatter structure. In fact, I've been a big proponent that the pyramid is going to look much more broader. We have hired more school graduates this year. We are hiring more school graduates this year than last year. Last year, we had 20,000. The year before, we had 12,000. So this is a more profitable business if you can push the pyramid broader. New roles and new job families, which you're going to hear from Kathy and lots of jobs on the front and lots of jobs on the back while AI is in the middle.

There is a ton of job families and jobs, which we have now established in this flow cycle. Business process reimagination and reinvention is the foundational step. We have a basis methodology, which is driven by our consulting teams, trademarked by us for transforming existing flows. We spoke about context engineering, the ability to put the work graphs, the guardrails, the sensors, the tribal knowledge and capture that, assembling that context is the new code.

Somebody asked me this question, what does this new frontier engineer do? And I'm going to come to that in a bit, and I'm going to talk about it. And that's an important part of this process. Our partnerships are now layered with frontier models, more AI native companies in addition to the incumbent SaaS and security companies and the hyperscalers we work with. One of the craft Cognizant built over the years is to take any new capability and build it at scale.

We don't buy capacity. We build at scale. If at all, we buy, we buy it for beachheading it. What is a frontier engineer and a frontier operator? I mean the market a little bit gets confused and they kind of use the forward deployed engineer as a proxy to it. The forward deployed engineer was built at a time when the world problem was everything in a company is broken, I'm going to stitch everything together. Bring it together on an ontology.

I'm going to bring a black box and an engineer who will fix it. That isn't the world problem we're trying to solve today. The world problem we are trying to solve today is you have flows in a company, you have to open the flows, audit it, do the wells, integrate it back and reinvent the flows and deploy it. A frontier engineer is somebody who brings that machine could be any frontier model. It could be different frontier models for different tasks.

That's the craft we have. And you reinvent those flows. So you need to have capability at the intersection of technology and the domain. A frontier operator, some of our clients will say, you know what, if this has digital labor and human labor, I'm not going to run with this. My core mission, if I'm a health care company, is to underwrite my insurance and to acquire clients, you manage the operations.

We need a BPO+++, which can use digital labor and human labor together to deliver those outcomes. That is the kind of capability we want to build, and we want to build it at scale, a frontier engineer and a frontier operator. We're going to build it at scale so that all our clients can actually thrive on this. Flatter, leaner, modulars, pods, non-STEM disciplines because we are now getting into other areas.

We have new skills, new ways of skilling and new ways of learning. You're going to hear from our Chief Learning Officer about our Skillspring program. The Skillspring training platform, which is actually a micro-personalized at your pace and it navigates to the finishing line.

We have an AI fluency meter for the 350,000 employees. We are flipping this and now we are actually presenting it to our clients, embedding it into the work we do. In fact, we believe learning is a part of our infrastructure stack, our AI infrastructure stack. So that's the change we are making inside the company as well as the transformation we do with our clients.

Our platforms business is a new business unit. We have a heritage with TriZetto. We have 200 million lives on TriZetto. We want to make that a 60-rule business. A 60-rule is something which margins plus growth kind of gets there. That's our aspiration. We sense what we need from our labs, our award-winning AI labs, you're going to hear from my colleague. We build, buy, partner, invest.

We have a new investment arm. And we think we can do it in 2 swim lanes, engineering to increase velocity and cost of deployment of AI and business operations. So business platforms like TriZetto, either we build, either we buy, we invest or we partner because there are -- I mean, look, software companies have to reforge as well because they have to own outcomes.

So they would be looking for partners who can actually take the end-to-end responsibility. We have physical AI, which we're talking -- which my colleague is going to talk about, and we're building a platform called Intelligence Spine. So here is the shifting economics from labor to outcomes, token optimization, AI-infused rate cards. A few of my clients are already using AI-infused rate cards.

A0 being all human effort, A1 being human effort audited by machines, A2 being machine effort audited by humans, A3 being autonomous. So we are embedding human and digital labor in a rate card because some of our clients are saying, I can't fixed price this. I can't do managed services. I can't give it on outcomes because we have pods between your teams and my teams. So we have AI-infused rate cards. I have clients who are already using it.

And we are already underwriting outcomes in TriZetto because we do it per person, per member per month for TriZetto. So in summary, the bridge from AI capability to production value needs a company like Cognizant more than ever before, the opportunity is $6 trillion of market versus $1 trillion of market for systems to $1 trillion to $6 trillion of systems plus people plus digital labor. And the capability gap, the production value to capability gap, you need to be an AI builder and Cognizant is poised to be that AI builder that helps enterprises bridge that gap between AI capability and production value.

So thank you so much for listening to me today. What I'm going to do in my next section of this today is to invite a partner of ours. The company is called Workfabric. Rohan, the Co-Founder, is going to be here. We've been working together on context engineering since 2024. We have done some pioneering work for our clients. These are a bunch of researchers from Harvard University and Carnegie Mellon who put this up, and we worked together.

We have also applied this to our own selves. And he's going to tell you a little bit of context engineering and also show you a little bit about our sales transformation and how we are using context engineering to generate a pipeline. And he's also going to show you about our customer success program, which is about anything which happens in a client, we sense it and we tell our clients before our clients tell us.

So these days, if any of my clients wants to call me up for a potential friction, I should be able to tell them before they tell me. And I should be able to tell them how I'm addressing it. So we are able to do that with the power of context engineering, and he's going to show you some of it. And we have powered our sales opportunity management system using context engineering. So Rohan, over to you.

All right. While this -- we just wait for this to load up, the context to what I will show you is, as Ravi mentioned, in October 2024, Ravi and I met and at the time, the predominant world view was that, as you saw in Ravi's graph, model capabilities will keep increasing, model intelligence will increase. But the problem with that is models don't really understand organizations, the nuances, the specificities of organizations.

So how do you actually solve for that? Because unless you solve for that, models will never speak the language of the organization and produce value in the context of the organization. At the time, this was not a popular idea. It was certainly not an accepted idea. There were just 2 guys in a room who believed this idea. Zoom to a few months later, we had an actual working deployment. We had built technology that could learn from how people worked in organizations that could extract tribal knowledge inside organizations.

We had deployed this in a Fortune 500 company. We saw some results and the transformation in terms of what agents could do when you grounded them in the reality of a company, in the context of a company was dramatic. The outcomes were dramatic. Ravi and I published our first article in Harvard Business Review in April of 2025. Insofar as we know, it is the first ever documented case of the value of context inside the enterprise. Since then, this whole space has taken off. You'll pretty much hear everybody talk of context engineering.

What I'm going to show you today, though, is a live demonstration of what Cognizant and our company are doing together with context engineering. I'm going to show you outcomes that at least as far as we know, have never been possible before nor have been demonstrated anywhere else, okay? All right. So with that, what you see here is a live system. This is a twin of various accounts inside Cognizant.

Obviously, for all the right reasons, we have had to strip out names and logos, but the data is very real, okay? And the use case here is as follows: -- any company, certainly Cognizant included or any company that is selling to another enterprise has a reasonably large surface area that interfaces or touches their customers. Sales talks to customers, delivery talks to customers, support talks to customers, finance talks to customers, et cetera, et cetera, et cetera, and so on.

But the intelligence about what is happening when you interface with that customer, what you're learning about that customer is not captured in any system of record. In fact, your CRM pretty much only captures what your salespeople put into it, right? But what about all of the other intelligence? Maybe there's a junior person in your organization who learns something very relevant about your customer that could have been useful to your salespeople, maybe your delivery person learns something and so on and so forth.

So there is this incredibly powerful intelligence about customers that organizations are always learning, but this intelligence is living in silos beyond the sales teams. We're historically used to thinking of having a record -- system of record like a CRM and the CRM will tell you the truth. Well, that truth is only good as whatever you put into it.

But what about what people are learning, the tribal knowledge, the tacit knowledge that you're learning about your customers? And so what we have done together with Cognizant, in our opinion, this is the first of its kind use case in sales is we have used context from within and across Cognizant to find net new revenue opportunities that no human could have ever found on their own, okay?

And what you see here is each salesperson in Cognizant, when they log in every day, they see for the accounts that they're mapped to leads that AI is suggesting that they should be chasing. And these leads are generated using context from across the entire organization versus only what is in sales. And therefore, by definition, humans can't find them. And for a moment, if you think about what I'm saying, it's really important because the most dominant use case of value for organizations from AI has been code generation or at least that's what's widely accepted.

But what we are talking about here together is this is direct ROI because we are producing net new pipeline. I'll spare you the specificities, but I'll give you a sense of what is possible. So here, as an example, it's telling the salesperson, you should go to Apex Auto, obviously, not a real name in this particular case, but the story is real and go pitch Cognizant's Flowsource for QA optimization.

And now when the salesperson clicks on this, what this is showing him or her is based on the context learned across different teams in Cognizant, there is the staffing levels for QA at this particular customer are under scrutiny.

Now when we double-click, there are more details because we are capturing this context from the lived reality of these teams from e-mails, from contracts, from documents, from conversations, et cetera, et cetera, and so on, of Cognizant's own people, while making sure we have privacy and customer sensitivity and all of these guardrails put in place, we are learning from the sales teams, from the delivery teams from people across the globe that there are signals that are telling us that in this particular customer, there may be some increasing oversight on QA staffing levels.

And the second point is, it looks like this customer has a mandate to reduce engineering costs by 15%. And the way we actually learned that was by listening or by learning from a couple of people who are spread across the globe who had picked up various bits of information that this was a stated objective for a customer. Some of it was in some ticket that was seen in Manila, some of it was seen in India, some of it was seen in the U.S. and so on and so forth.

And when we put these signals together, we proactively realize what the priority of this customer is, what is their pain point. And therefore, what can Cognizant proactively go do for this customer versus wait for them to come and ask you something. And so therefore, in this particular case, it's telling the salesperson, go to this particular customer and pitch. Here's a proactive pitch. The customer didn't ask for this, but go proactively pitch to them saying, "I think you may have this pain point. Here is how I can be of service and of value to you.

And then, of course, there's a briefing generator and e-mail to make it easier for the salesperson and so on. Now what this is essentially doing is it's producing a class of revenue opportunities that are synthesized by the living reality of what Cognizant's own employees are learning about their customers, right? Now later on, I'll leave -- Ravi will tell you there's a dramatic number because this is not a pilot. This is not a proof of concept. This is not an experiment. This is not a research project. This is a living, breathing system that is running on thousands of Cognizant's people, their teams, their entire global sales force producing this kind of an outcome.

I'll let Ravi give you the dramatic big number. Now the second quick thing is, all right. So now that we have context, you take an account, you have context of the account from people who touch the account from these different teams. And so Ravi mentioned, he wants to be able to go to a customer and say, here is a problem that I'm already aware of and we are fixing it versus wait for the customer to come and escalate you and say, there's a problem, I'm not happy.

So the way we address that using context is now we have built these mission-driven twins of these accounts. What does that mean? Well, so we take the context of all people who work on an account. We put that together, and we've created a twin of the account itself, using this context. And now we give each one of these account twins a mission.

So you're seeing -- again, these are real accounts, names have changed, obviously, but each one of them have different missions. So for example, there is an account where the mission is prevent revenue churn. There's another account where the mission is find more revenue that is likely to close in 120 days. And these missions are live.

So when Cognizant is not looking, it's supposed to be doing its job and finding new revenue that people may not be aware of or build a renewal strategy and so on and so forth. You can pretty much create any kind of mission. But I'm going to show you one particular account and the mission there. Again, the facts have changed very deliberately, but it will give you the flavor. So this one needs attention, okay?

And this is Cognizant proactively recognizing that it needs to pay attention to this account. So let's click on this. Again, remember, this account is collecting context globally from everyone servicing this account. So it turns out, for example, on May 11, this mission was accepted by this twin, all of this using context. In this account, they were doing -- Ravi earlier alluded to modernization and SAP migration. This account is doing SAP migration in multiple sites.

Now based on the context of the various teams, it looks like the Dallas site went fine. The Sydney site seems to go fine as well, slight anomaly in Riyadh, but that's okay. It's a holiday. But then there's a first sign or a signal that is telling us saying there may be an issue in Sao Paulo. And we picked that up because somebody within Cognizant learns something from the customer giving them some kind of feedback.

Usually, that would have just been lost or you would have waited until that accumulated or escalated. And eventually, the signals get stronger, and it looks like the deployment in Sao Paulo and to give you a sense, the context is from these various teams from the on-site team, the delivery teams, et cetera, et cetera, and so on. And there are various signals that are embedded in these teams that are telling us saying this particular site may have an issue.

Again, now post this, this account twin pulls context from the SAP practice twin. There's a twin even of the SAP practice in Cognizant saying, is what's happening in Sao Paulo based on its context, very different from what you usually see. And if it is, okay, it looks like we need to do something here, it suggests to the human, maybe you should pause this now. The human takes the decision. So ultimately, the humans are still in control. Pause this. And then the twin comes back and says, here are 2 people based on their own context who might be experts in helping resolve this situation.

Eventually, it also helps to draft an e-mail to the customer and so on. And then it creates a briefing for the CEO because in case there's an escalation or if the CEO wants to pick up the phone and call them saying, don't worry, we know there's an issue with the Sao Paulo deployment, but we will take care of it. It's a very detailed briefing.

All of this, the key point is powered by the tribal knowledge and the context of individuals inside Cognizant. Okay. And there's more -- obviously, more to see and more to show you. But hopefully, I've given you a sense and a flavor. You hear the term context engineering, but I wanted you to see -- we all wanted you to see what Cognizant is actually doing in production with real numbers and real deployments. And hopefully, this gives you a sense. And with that, I'll hand it back to Ravi.

So at this point of time, we roughly have $200 million of pipeline generated incrementally through this extraordinary effort of doing a sprawl on the systems, e-mails, meetings, chats, everything else and generating it. By the end of the year, we think this is going to be a $1 billion pipeline just by listening to conversations of our clients and extracting the tribal knowledge to create some insights.

In fact, there's another interesting use case, which Rohan didn't mention, which is about how do we staff programs. Today, we staff based on resumes and skills in a system. Every time we have to staff something, we just have to ask the context engine, tell us who in the company is doing similar work, who will fit the bill. It will find people around and say these are the 10 people who can fit the bill.

So we've done that as well. With this, what I'm going to now do is everything we spoke about, what does it do to our operating model, what does it do to our business? Jatin and Ravi Kiran are going to talk about it. And Jatin is our Chief Financial Officer. Ravi Kiran is our Chief Strategy Officer.

And before they come on stage, we're going to have one of our customers, LPL Financial, which is actually working with us to power the wealth management platform with Agentic to support their wealth advisers. So we're going to see the video and then get to Jatin.

[Presentation]

Okay. So hi. Good morning, everyone, and thank you for being here. Very happy to see so many of you being present to hear our story. The purpose of my session -- my and Ravi's session is really to speak about the journey for our customers from something that looks incredibly exciting from a technology potential standpoint to real value creation for themselves and therefore, their own shareholders, right?

I mean that is the end goal. And for that, you need 2 things. You need a great strategy, and I hope you heard that well from our CEO. And I'm going to speak about how does the company behind changes to align with this strategy and deliver that value to our end customer. We think this operating model has 5 legs, which are crucial. The first is how we sell. The second is what we sell. Third is how we deliver the new offering. And fourth, which I'm sure many of you are very interested in knowing, how do we monetize this expanded surface area?

And finally, how does the talent or the organization that all of you have seen, touched, understood for the last 3 decades changes and becomes an organization of future. These are the 5 elements of operating model, which are very crucial for the execution, which one of our customers just spoke about, that execution has to come from the operating model once you know what the blueprint is. We'll start with GTM first, and I'll request Ravi to speak about it.

I think from a go-to-market and the way we take to clients in terms of how we sell our services, obviously, every technology shift that has ever happened always has led to IT services players like us helping clients through their journey. That has always been the case.

And when you look back at the cloud era as an example, I think we've been able to help our clients with the entire lift and shift to the cloud, build native cloud applications, do the integration, a lot of the infrastructure that was needed underneath cloud was kind of something that we have navigated our -- most of our clients through that journey.

I think with AI, it's going to be no different. I think AI, in fact, what Ravi alluded to, being an AI builder and the fact that there will be more and more bespoke AI capabilities that have to be built needs a massive uplift from service players like us. And I think it's also the complexity of bringing and orchestrating a whole bunch of ecosystem players. So you're looking at frontier model players, you're looking at cloud, you're looking at security, and now you're looking at infrastructure, a lot of things that have to come together.

So in that sense, I think it is going to be a massive need for players like us to step up the game and take it to clients. One of the things, obviously, from an operational standpoint that we have enabled is creating an AI market unit, making sure that we're bringing the strength of the organization and bringing it as a single interface to our clients. So it's something that's critical to taking it forward.

The second is deepening our partnerships with many of our partners, the frontier model players. So one of them, some of the security players that Ravi alluded to as well. There's a lot of that in terms of how we kind of bring together a far more integrated model that we can take to clients where there's more of token and inference models that can embed into our go-to-market strategy itself. And the last thing is obviously about -- you've seen the demo from Rohan.

I think it is a phenomenal demonstration of how we have leveraged context engineering to kind of build our sales capabilities. Our entire sales engine is enabled with these context sensors and the capability behind it. And to me, I think this has only amplified the sales capability to take it to the next level. And there's more -- obviously, we are seeing a lot of benefits that are coming out of it. But to me, this is in a quick summary, the operationalization of some of our GTM activities.

I want to speak about what we sell. And Ravi had a whole slide around vector 2, vector 3. And I would just double-click on 1 or 2 of them, how this is changing. And again, I mean, we all have been in the industry for more than 2 decades, almost everybody in this room. You know this, that whenever we provision something, that goes first and security follows. And then you put a wrapper of security around it. And if you are really smart, you think about security even as you're developing that.

Now look at how AI is changing this. And I'm giving this example because when we speak, the first thought that comes to our mind is, but AI is deflationary for services. But I'm telling you an example of a service that is enforced because the AI exists, which is Mythos. You start first with services, with security and look at this, the entire paradigm of 3 decades has changed. Security was following now security leads.

Security at the front tells you that these are the vulnerability, which is sitting in your organization, now go fix it. So this is -- this didn't exist 6 months back. This exists today. I'll tell you one thing else. If you look at in security, every time you put an element in your network or element in your outcome that you're generating, you need an extra element, be database network, applications, cloud, you know that your earlier security posture has completely changed.

Now you need to resecure it with a new element of AI, which is now helping you deliver services. It's very easy to say, okay, let's put AI and get a better outcome. But you are not thinking what you know differently on security. So just on a business of security, you have now 2 very large opportunities. The first, what we call is AI for security. And second is secured AI. By itself, large momentous opportunity, Y2K opportunity for us.

Now if you look at the second is really operations and tech. Traditionally, operations and tech have run parallelly. IT services companies have large operations practice. But even when customers make a decision, they said, okay, let's give operations to player A, tech to player B. Now you have agents which are sitting in the middle where somebody can do what client is doing today, which is teaching a tech and ops together and create an outcome. That's the whole new opportunity.

One simple example of a new service, which has come because of AI is this training the agents, which we do in ops, which was not present again a year back. So we feel very excited. The composition of services that we are going to sell is going to change materially in times to come. And even our inorganic investments are all aligned on that.

3Cloud that we acquired in the beginning of this year, the leadership here, I would encourage you to speak with them, but they are really focused on supporting organization in their cloud journey as they leverage AI. Astria really focused on creating -- on helping create the data center -- the technical help to create data center, which we all know is a big investment in today's time. So we feel very energized about the new offering or the new pools that we can tap with AI.

Okay. The third dimension of our strategy and the operating model is about the platform shift. Unlike the past, where it was all about skills and platforms, we are now shifting into an era where it's going to be skills agents. You're going to have a fairly large Agentic layer and AI-infused platforms that's going to coexist going forward. So in that sense, I think a clear shift towards platforms that are going to be more outcome oriented, they're going to actually start measuring real value.

And one of the operating principles that we have used behind it is to start looking at the set of platforms that we have. And we've consolidated all the platforms into one unit. That was one of the first things that we have done. The second is obviously, how do we look at AI enabling the entire, even going to the extent of making them AI native and coming up with new monetization models so that the platforms can actually deliver value as part of the offerings itself to our clients.

And in a sense, I think the entire platform approach is something that we think is also important from a nonlinearity standpoint. Eventually, we want to get to a stage where the workforce is complemented with the agents and the platform so that there's a fair bit of nonlinear economics that we can derive out of it.

The monetization is crucial. Ravi spoke about headless TriZetto. Now imagine, I mean, we always have sold TriZetto in a per member per month basis. And that's a real constraint as only so many new users you can get in organization. Now that IP is exposed to the whole Agentic workforce of the world. Anybody can come and tap that IP, and we will have to monetize -- we have an opportunity to monetize that interface.

We spoke about A0 to A3 inference building inside the way we sell. And you could argue that inference could be margin dilutive. And I would say no. The reason for that is that, as Ravi mentioned, there is a secret sauce about how we use inference. Just the way we have a pyramid of human skills on left side, we'll have to build a pyramid of inference on right side. And then we leverage that pyramid to build the best outcome for our customers.

And that's the secret sauce that we can go with. So now traditionally, we have sold skills. Now we sell skills plus inference. That's a massively large opportunity that we never had before, and that's a $6 trillion that Ravi spoke about. And this is not something tomorrow. This is even as we speak, not a $5 million deal, a $6 million deal. These are large deals that we are submitting, which is built up of skills plus inference, and this is the outcome that we are generating.

We talked about how we sell, what we sell, how we deliver and how we commercialize. Now bringing all these things together, obviously, the foundation to all this is the workforce. How do we retool and reskill the workforce. That's going to be one of the foundational shifts that we are doing. I think Ravi talked about the new roles and the new career architecture that we have launched, the Frontier Certified Engineer and the Frontier Business Operator.

These are new roles that are evolving for this new AI era, where they're going to be a deterministic and a probabilistic ecosystem that's going to coexist. And that's going to be a big part of the workforce re-skilling that we are currently underway. Obviously, when you look at the way AI is going to impact a lot of our clients, the pyramid structure that we currently have today is almost no longer relevant.

I think we're going to be a much flatter organization. It's going to be a much wider organization. We're going to have a lot more agents at the bottom, a lot more interdisciplinary skill sets that are needed for the future. So in that sense, the middle layer essentially will be doing more of a player coach kind of thing that kind of Ravi alluded to earlier.

So this is a fundamental shift that we are doing. And the AI fluency is something that is a critical part of our strategy as well in terms of bringing together not just how you get trained, the kind of project exposure that you have, the kind of whiteboarding and other related hackathons that you are part of in various forums. How do you bring all these together to provide a fluency score for every individual?

And eventually, how do we take the fluency score concept to our clients as well. So how do you make the clients also kind of -- so this is where we are headed from a workforce transformation standpoint.

If you want to really take away 3 things that keep us driven every day on our operating model, this is the convergence of tech and ops and the new opportunity that comes with it. Blended effort. I mean, we no longer think in man months, we think about the total blended efforts of skills plus inference plus platform.

And finally, these 2 combined gives us the larger commercial area and how are we monetizing it because this is really the key of keeping 350,000 employees grounded on where we want to go and how we get there. And there are -- of course, there are KPIs. Some of them are old world KPIs, which will never go back to like 9 months growth.

Some of them are transient KPIs like token consumption. I mean, we were very excited about it 6 months back. Now we think it's no longer relevant. What I should measure is the throughput that I generate out of tokens. And some of them would be in external domain like revenue per employee, operating profit per employee, machine-assisted code and some of them will remain internal as we transform the company.

They will be relevant for a period of time and then they won't be relevant anymore. This is a quick summary from Ravi and me on our operating model. Thank you very much again for coming here. People who are listening our webcast, thank you very much for joining us. Our IR team and our media team is on standby if you have any follow-up questions, but thank you very much.

Thank you, Jatin and Ravi, for painting such a compelling picture of how Cognizant is really poised to be the AI builder that bridges the gap between the sheer velocity of AI and really the ROI that all of these organizations need to be reaping. Good morning, everyone. With that, it ends our livecast portion of today's event.

My name is Antonella Bonanni and I'm Chief Marketing Officer for the Americas at Cognizant. I'm thrilled to welcome you all here today, and I'll be your MC for the rest of the day. So before we dive into the agenda, just a few notes for some quick housekeeping. If you could please refrain from taking any photos of the recordings or the presenters or presentations. We're going to provide a packet of information for you post event along with photos of the event that you'll be able to leverage.

So with that, let's take a look at the agenda. If you could flash that up, please, quickly. You'll see that we have a robust set of Cognizant executive leadership team here to really walk you through the various dimensions of what we really -- day-to-day is first and foremost on our minds. You also see we have a nice lunch break before we dive into the rest of the day.

And before I introduce the first session of the next portion of today's festivities, we really wanted to have a listen from one of our client organizations CEOs. She's going to really echo the reality that Ravi Kumar painted this morning with his remarks. So with that, we'll play the video.

[Presentation]

[Audio Gap]

Starting to see unlock of vulnerabilities security-led with Mythos or 5.5 GPT, where you could do vulnerability discovery at machine speed, but you can equally remediate, refactor and patch at machine speed. And that opportunity is almost like a Y2K moment for us because as you open the lid, you don't just look at one thing. You look at brittle code, you look at design deficiencies. You do a whole lot of refactoring. So the ability to take your old stuff, pull in new ways to do it, underwrite that capital for the new things.

Now new things which I'm talking about is a completely new terrain, which is primarily applying software on operations of companies. Now historically, portions of those operations, which were deterministic in nature, were done by classical software. Portions of the things which were not deterministic, which had judgment, which had -- which could not be codified, we put it into human effort.

Just an example, the health care platforms in Cognizant have a flow-through of $500 billion of spend flowing through our platforms. We do health care operations for our clients. Now I could agentify another $4 trillion of work, which was administrative in nature -- remember, health care has 20 million people working, only 5 million to 6 million people do care, 14 million to 15 million people do administrative work. I could agentify it.

Now that doesn't need big discretionary spend. You just have to front-load it. You have to transition human labor to digital labor. It is self-funded. So you could take the steps even in an uncertain situation -- this is what I do with clients. Yesterday I was in my health care conference, we had 1,000 clients there. And the ability to take some of that and put it back into care I think is a phenomenal example.

So in summary, the framing I have in my mind is the AI capability is absolutely moving at a rapid pace. The bridge to production value has a big gap. The more the capability is, the more the production value gap is. And what does production value mean? This is a contextual technology. Production value means you have to ground the technology into what clients want their work to be transferred to AI, which means workflows, controls, guardrails, trust layers, evals, context and the tribal knowledge, all of that put together.

The more the technology is evolving, the more the technology is advancing, the more is the gap to enterprise value. Not for software engineering. Software engineering is already mainstream. I'm talking about operations of companies.

I mean if $1 trillion were spent in the last 1 year on infrastructure build-out, another $2 trillion to $3 trillion is going to be spent next 2 to 3 years, all of that build-out will only stack up if you are able to apply it to operations of companies. And the bridge to enterprise value, the bridge to enterprise production value, I think, is a big gap.

So clients are asking me, how do we actually bridge that gap? So that conversation is happening now. So therefore, I'm actually much more optimistic about how AI is a tailwind to our industry. You need a ton of work to be done to get to that production value.

And the production value will come from agentifying finance operations, legal operations, HR operations, customer service, and vertical operations like health care, mortgage operations, a whole bunch of things, which was a human endeavor. It's a different topic on a different day to talk about what will be the next endeavor, but all of that we have a struggle between -- interoperability between human and digital labor and building those flows in companies.

Yes. No, I like the gap discussion, being on the bridge, and I think -- and AI being a tailwind. I think these are all important messages. But investors...

More the capability, more the tailwind.

So investors are, of course, looking at the financials and looking at the performance, right? And the news cycle is so intense, Ravi, and it's difficult to translate, right, what's cyclical versus secular. I'm curious, just to get it out of the way, I know you just set your outlook just, what, 3 weeks ago, and there's been so much change since -- and we'll talk about some of it. Has that changed your confidence in the outlook and what you see in the near term here?

I mean look, 2 to 3 weeks is too less to really change the paradigm. But the one thing I can certainly tell you, everything I said so far is no longer conversations yet to happen. The conversation is happening now. I already have a large mainframe deal where I'm using Claude and transitioning work to AWS cloud. I have at least 5 to 6 SAP migrations happening. I have 10 to 12 discoveries happening on Mythos and 5.5 GPT. So those conversations are accelerating.

I would say, if you go back to the last 2 years, the industry was layering consolidation, an expansive spend related to consolidation, as a growth lever. That is right at the base I'm layering in these new value pools of old things done in new ways, which is unlocking mainframes, unlocking SAP migrations, unlocking vulnerabilities on landscapes and layering it further.

Once I layer in the spend on the operations, which is also self-funded, it doesn't need so much discretionary because you are actually eliminating work versus creating more work, once the unlock on growth happens, you are going to use the same set of people to create more throughput. You're going to see new products and new services powered by AI.

So if you put those -- if you put that layering in, you will see an inflection point at some point of time. So this is not about -- my framing last year was, what does AI do and what does humans and tech services company do? My framing has completely changed. My framing now is the capability is right up there and it is further going up. The production value is here, and the gap is huge. That gap needs a bridge. And that bridge will come from companies like us.

If you believe that the capability will not actually impact operations, then you almost have to question why the infrastructure build needs to be that much. So that's the way I'm framing it. I mean it's a question of time this inflection point has to happen.

I mean Financial Services for Cognizant is already a double-digit growth. My BPO operations, which is where the expansive opportunity is, it has been on double digit for 2 years in a row. Infrastructure Services, because there's a sprawl of infrastructure, that is close to double-digit growth. Software engineering is deflationary. The expansion attached to software engineering is taking off now. So I mean, we just have to keep layering this and we'll be back to industry-leading growth.

I mean we're already on the winner's circle and I'm less interested anymore about being on the top of the heap. I just want to take this to breakaway growth. And that is my new aspiration. I mean being on the top of the heap is no longer good if you're on the top of the heap. You should actually look for breakaway growth.

Yes. No, I'm glad you said that. And look, getting back to the winner's circle, I know it's not easy and that was a lot of your doing from an execution standpoint. So you should be credited for that. But you're right, focusing on that is missing the bigger picture.

And I'm asking -- the reason why I asked, right, because the news cycle is changing so much. We're going to hear from OpenAI tomorrow. And everyone is asking me to ask you, so I'm going to ask it here, right? With OpenAI and their DeployCo, and then Anthropic similarly working with these investment partners to build out their own deployment or services capability, what does that mean? Are you talking about this group that [ will cause that ], the change in the competitive landscape in your mind? Or is it more of the same?

No. It doesn't. Actually, on the contrary, it reinforces the fact that there is a gap between production value and the capability of those models. We underestimate the heterogeneity of enterprises, we underestimate the fact that this is a contextual science. Everything which was supposed to be deterministic has already been in classical software. This is deterministic, it has to be grounded, it has to be harnessed.

So on the contrary, it reinforces the fact that we need that bridge. You can argue whether that bridge will be companies like us or it will be new companies coming into picture. I don't think those deployment companies have been built for scale. I don't think it is built to make -- monetize the bridge.

It has been built to get access to the distribution network, which is needed. And that distribution access always was the reason when new technology came into picture. When enterprise software came into picture, we had Oracle Consulting in the mix. When SaaS software came into picture, we had professional services supporting SaaS software. When cloud -- the embrace of cloud happened, we had professional services coming into picture. But it never took off at scale to build what is needed for the Global 2000.

So I see this as an endorsement of what we do. I see this as an endorsement that this bridge is needed. Do we need to do it differently? Yes.

I mean everybody is now talking about a forward deployment engineer. Forward deployment engineer was the perfect thing in the last few years when the worldview was everything was broken in an enterprise, you have to bring everything together, you go with the machine and you go with this forward engineer and you tie it together. The worldview is no longer about everything is broken. The worldview is you want to take this machine, which has digital labor, along with you, and you want it to be integrated and you want it to deliver an outcome.

So we have to reforge our first principles. Our first principles have to be reforged where we go from a system integrator to an AI builder. I've mentioned this. We go from a pyramid of talent to integrated interdisciplinary talent with human and digital labor. We go from a services company to a platform-as-a-services company. We go from managing project outcomes to managing operational outcomes for enterprises.

So the roles we need now is frontier engineers and frontier operators. Frontier engineers being people who can engineer equivalent to in a way of forward engineer, the machine at that time was ontology with machine learning. Now the machine is generative AI.

You also need frontier operators. What I mean by frontier operators are, if clients are going to run operations with us, some clients are actually going to say, "You run those operations for me with human and digital labor." And that is not what frontier engineer is about. Frontier engineers, "I engineer this product."

I'll give you an example. I have 5 banks who are doing Know Your Customer with me. They don't want to outsource that function. What they're essentially saying is "We love what you're saying on agentic. We want to agentify. Give me a fixed price proposal for agentifying Know Your Customer."

There's another set of clients who've come and told me, "Here are the stuff which you do in health care today on your TriZetto platform. There is a ton of labor sitting outside. It could be codified. Please identify, run those operations for me." So I need a BPO professional who can work with digital labor and have an integrated digital and human labor and actually deliver an outcome. That is the new-age BPO we are actually building, and my BPO business is actually at double-digit growth.

So you need frontier engineers for things which you want to engineer for your clients, agentic. You need frontier operators for things which you want to actually own the outcomes, operational outcomes.

So with TriZetto, we are going from -- we went from perpetual license to subscription license, to BPaaS. Now we are going -- we have started to work on per member per month to manage lives. It doesn't matter what transaction. In fact, in an ideal world, you shouldn't have transactions, you shouldn't have claims. It should be a claim-less world. So just manage per person per month, here is the menu card. So we have to be a platforms company.

Either we have to build our platforms or we have to work with third-party SaaS companies and integrate their platforms into our services. Because clients are neither going to look for SaaS software on its own in some places, neither are they going to look for service. They're going to look for AI outcomes.

So what else has changed in the last few weeks? Tokenization has become a huge thesis for me. And I'll tell you, in my earnings, I spoke about AI-enabled rate cards. A0 being human effort, A1 being human effort validated by machines, A2 being machine effort validated by humans, A3 being autonomous.

And now clients are saying as you go from A0 to A3, the premium on the rates will go up, but the number of units will go down because there's machines attached to it. Now clients have started to say, "Wait a minute, you're going to take the human effort accountability, I'm going to take the machine effort, and I don't know how to do this. So take this over and run," which then means I should build a harness around tokenization. Why? Because I could do it with hundreds of clients, so I should be able to do the same things in a compounding way.

That's why clients came to us. Clients said, "Oh, you do this work well. You've done it with 50 clients. Give me the people who have done this before, show me references, I'll give you the business." That's how clients feel.

Now tokenization could be a new function value for IT services. So we have metered inside the company projects, people and the kind of work we do. That metering leads to institutional harness. Earlier, it was at the minds of people and repeatable artifacts. Now it is in the harness. I should be able to deliver your digital labor more reliably, cheaper, cheaper is important, till token costs are up, and more predictably, which then means I take a high risk on tokenization, higher risk on inference costs on tokens, if I'm doing operations. If I'm doing development work, the token costs are less risky. If I'm doing inference work, I should be able to.

So we have now started to build a craft on tokenization. And this is a discussion many of my clients are doing now. They're saying, "My bills are going up, would you be able to take this over? We use these tokens to process invoices or process claims. For each of them, we have a way to spec it, we have a way to size it and we have a way to price it." So effectively, that's a new moat. And it will come from the community knowledge of clients because clients actually -- if you do it at 30 places or 50 places or 100 places, you'll be able to do it better than your clients. So that's the craft we are building.

Is it completely efficient? Not yet. Am I going to falter? Maybe yes. But I'm going to build a moat, which will help this tokenization process to be more scientific, versus today you trying to measure tokenization based on consumption versus measuring on value.

This tokenomics thing -- thank you for going through all of that, right? But we get this question a lot, right? I mean creating a new model and then pricing a new model, and you're going back and forth with the clients and you're learning. I appreciate that you have a lot of scale and you're accumulating all of these tokens. You can price it differently than consuming it individually. But what's the risk of something going wrong, Ravi, in terms of -- just like taking it back to the old fixed cost.

Yes, yes. Absolutely. Absolutely. In software engineering, the risk is low because...

[ Template with that ].

Also you are developing something, you're not running it. The cost of tokenization is a higher risk in inference and lower risk in developing. If you're running operations of companies, you have to be precise on the tasks you want to do for a client and you should be able to spec the job. So you have to pick your swim lanes where you want to. I mean you could be blind-spotted by things which have high inference costs.

So I'm not going to run wild on this model. I want to run in a controlled way to figure out which swim lanes we can precisely size it and which swim lanes will have throughput so that we can build patterns and, therefore, the compounding effect. I mean the compounding effect is similar to autonomous cars. Autonomous cars drive better than you not because they have your learning. They do better than you because they have community learning. So wherever there is a repeatable rinse-and-repeat template, you should be able to do it.

Now is this going to be a long-lasting moat? I don't know. If the cost of tokens is so low that it's like a utility, then cost is not going to be a driver. It has to be driven by velocity and, because it's a contextual science, it's also driven by reliability and predictability.

So adoption-wise, how quickly do you think you'll learn and be able to evolve and commercialize this model? Is it months? Is it quarters? When should we be asking you the right questions on, hey, you figured it out? Forget about the moat, more the adoption.

I would say the ones which have been historically outsourced, like health care operations, mortgage operations, customer service, F&A functions, are easy or relatively easy. The ones which have been historically not done, like legal operations have not been outsourced to companies like us, HR operations have not been -- that will be harder. The vertical ones are a longer runway. The horizontal ones have a lesser opportunity, if I may. Because the vertical ones actually have the most administrative labor which is -- which can be actually shrunk.

I mean I told you about health care just as an example, 20 million people work in health care, 5 million to 6 million people really do care, 14 million, 15 million people do other things. And it's a high-churn industry, so it's not like employability at -- on a sustainable basis.

So I think we have to pick the right ones. Health care is certainly one I'm picking. In fact, I want to make sure that the hustle is around TriZetto. So the rules engine of TriZetto, we want that to be the real moat from which agents are accessing rules and guardrails and everything else. So we want to make sure that that is where it is.

If I have to pick a few more, Tien-Tsin, just at a high level, wherever classical software did it in a less effective way, like billing systems, recordkeeping systems in life insurance or mortgage operations. I mean these are the places. These are not the best-looking things in classical software. They potentially become the best-looking things for AI.

Okay. So thinking, as we maybe hopefully get you back here next year, thinking about this developing and your reforge principles, how ahead of the curve do you think you are on this, Ravi? Because it feels like there's some margin accretion potential and the usual financial questions we'd ask. But just thinking about how far ahead you are versus the peer group, how confident do you feel around that?

Look, I'm layering in consolidation of work, which is giving me some runway. I'm layering in these new value pools of old things done in new ways, which is giving me some additional incremental growth. Now I'm layering in new things in new ways. New things in new ways is also operations, self-funded. Once we see the macro improving, you will see new products, new services coming in.

I think last time I spoke to you, we spoke about a digital nurse, which is a new thing. It's not a labor pool existing. A digital nurse for people who are doing dialysis treatments in their homes. We built something like that. We're building something on wealth management. So we're building a new -- those have not fully taken off except Financial Services because of discretionary. So that layering will certainly happen.

So in a year is what you're saying, right? In a year from now, I think what you have to judge companies in this sector, including us, is are we -- I mean, it's hard to say what is AI revenues and what is not AI revenues. So it's hard because there's no real baseline. Revenue per person, margin per person is an important metric.

And I would say revenue per person and margin per person because I've changed my pyramid mix with our Leap program. What I'm really saying is the pyramid is broader and the pyramid is shorter. So I'm delayering the nodes where there is administrative work and I'm only having player-coaches in the middle. And I'm hiring significantly more number of people at the bottom, higher than last year, and last year we hired higher than the year before.

So that whole thing will give me margin per person higher, but it will not give me a revenue per person higher because I'm broadening that. So I want to create the right balance. Right now, we've got a runway on revenue per person, we've got a runway on margin per person. Margin per person has got a higher bump than the revenue per person, in my case, because we're broadening the pyramid. But both should get a bump. That's my sense.

Now what we have achieved so far, is it good enough? It is, I would say, it can be better. Revenue per person and margin per person in the TTM basis has just about got to double digits. So we have to expand that. The platform play will allow us to increase the revenue per person and the margin per person, because platforms are an important part of my thesis. We have set up a new platforms group, dedicated. Some we want to build, some we want to buy. Some we are going to invest. We put up a venture arm, we announced a few weeks ago, and some we're going to partner with third parties.

I think the throughput from the 3 frontier model companies, Anthropic, OpenAI and Gemini, all 3 are going to be super important because that's the machine we are carrying now. I think you should also -- I don't know how to judge this on a common scale, but the number of frontier engineers and the number of frontier operators we've been developing. I mean we are a company and we are an industry which has not built human capital based on buying stuff. We've built it based on building it.

And this is a new -- we are getting to a new era where a new set of companies and a new kind of people who will express in those companies will work, and those new set of people are people who can actually take these frontier models and deliver operations for companies. So I am measuring it on certifications, which is an external metric. That is something I'm very keen to pursue.

And I think for our scale, will be super -- it -- that's the kind of capacity needed to bridge capability to production value at scale, at much better economics. At much better economics. So I mean if you put all of this together, you should start to -- we should start to go back. I don't know when. But we should start to go back to growth rates which make the industry attractive.

Yes. No, the plan makes sense and it's going to be fun to track for sure. I have to ask one follow-up to that though. Thinking about the capital intensity of doing some of this. I know you're acquiring Astreya and getting into managed services infrastructure. So getting to where you want to be, Ravi, is there a need to do more inorganic to get to the right place? How do you balance that?

Yes. So we did a buyback through the morning, for $1 billion. We increased it by $1 billion. We had $1 billion of buyback in January. We felt like that was a great way to show confidence and great way to back our story.

I think the M&A we do should fall into this thesis of platforms are operations. And we'll be very purposeful in doing that. Astreya has not closed yet, but one of the reasons why Astreya made phenomenal sense to us is the infrastructure sprawl, which is going to happen, we are already close to double-digit growth in infrastructure, we felt it is an extraordinary opportunity for workplace services, data services, data center services and network services. That's what the company does. And it doesn't do it on human effort; it does it on a user basis.

So they do it on a per user basis and they do autonomous infrastructure in workplaces. So it's a complementary capability, platform-led, and a new outcome-based model because it is not based on human effort, it's based on the number of users. So it fell into our thesis, and that's what we did.

I mean would we be opportunistic in places where we have lower presence? For example, I have less presence in Asia Pacific, I have less presence in Europe. Of course, we'll do it. But my general strategic intent on M&A is to layer platforms and operations because that's where the bridge to enterprise value, the bridge to enterprise production value is very high. So that's been -- that's how I look at M&A.

Okay. No, that makes sense. So we're just about out of time, Ravi, I wish we had more. But just thinking about piecing this all together, and hopefully we'll have you back next year and I'll ask it again, but what outcomes do you think will be most obvious that you're pushing towards that will come back and be like, "Oh, wow, your progress and your reforge principles are working?"

Certainly, revenue per person, margin per person is one metric to watch. I would say more fixed price, more outcome-based pricing. We have 50% fixed price, 10% outcome-based roughly. We moved -- we flipped the whole thing. We were -- we almost added 10% into fixed price in the last 3 years from 2023 to now. Fixed price because the engineering on AI where people are not willing to give you the operations will potentially go fixed price.

By the way, I have to make that distinction there that the AI-infused rate cards is also new stuff. It might show up on time and material. But it is new stuff because I've been fusing digital labor and I've put this new stack. And we are the only one who are actually propagating this to our clients. But that will be another metric to say we are progressing that way.

The work we do with our platforms group is a very important change in our reforge first principles. The work we can do on outcomes is a combination of fixed price and transaction-based or outcome-based pricing. These are the reforge first principles on which you should see whether we're progressing in the right direction with the right velocity.

Okay. Good. No, the fixed price piece, I think, is important. No, I appreciate you going through all of that. You're pushing through a lot of change, which I know isn't easy and is forcing us to study. We need to come up with a way to phrase the forward operators. We need an acronym for that. But I appreciate all this, Ravi.Thank you so much.

Thank you.

Thanks for the time.

Thank you so much for hosting me.

Greetings, and welcome to the Cognizant Q1 2026 Earnings Conference Call. [Operator Instructions] As a reminder, this conference is being recorded. [Operator Instructions] it's now my pleasure to turn the call over to Tyler Scott, Senior Vice President, Investor Relations. Tyler, please go ahead.

Thank you, operator, and good morning, everyone. Welcome to Cognizant's First Quarter 2026 Earnings Call. I'm joined today by Ravi Kumar, Chief Executive Officer; and Jatin Dalal, Chief Financial Officer.

By now, you should have received a copy of the earnings release and investor supplement. If you have not, copies are available on our website, cognizant.com. Before I begin, I would like to remind you that some of the comments made on today's call and some of the responses to your questions may contain forward-looking statements. These statements are subject to the risks and uncertainties as described in the company's earnings release and other filings with the SEC. Additionally, during our call today, we will reference certain non-GAAP financial measures that we believe provide useful information for our investors.

Reconciliations of non-GAAP financial measures where appropriate to the corresponding GAAP measures can be found in the company's earnings release and other filings with the SEC.

With that, over to you, Ravi.

Thank you, Tyler. Good morning, everyone. Thank you for joining us. We delivered a solid first quarter with revenue growth in the upper half of our guidance range, expanded adjusting operating margin and strong bookings growth. I believe our work to become the world's permanent AI builder is resonating, demonstrated by our first quarter performance.

Looking at the quarter's highlights. Revenue grew 3.9% year-over-year in constant currency, led by strong performance in North America and driven in part by the ramp of recently won large deals. From a segment standpoint, Financial Services grew more than 10% year-over-year in constant currency, driven by strong demand across banking and insurance clients. Q1 bookings grew 21% year-over-year. We signed 7 large deals with TCV of $100 million or greater, including 1 mega deal valued at more than $500 million.

Importantly, we continue to drive profitable growth as adjusted operating margin expanded year-over-year for the fifth straight quarter. Adjusted EPS of about 14% year-over-year was ahead of revenue growth. And we just announced a definitive agreement to acquire [ Atria ], a global IT managed services provider and a specialist in AI infrastructure build-out with deep expertise in managing data center infrastructure, enterprise networks and digital workplace technology.

Upon closing, we believe [indiscernible] will add a critical layer to our AI Builder technology stack. We achieved these results against a softening demand environment. Market conditions have become more complex since the start of the year, and we expect the impact from heightened macroeconomic uncertainty to persist in the near term. However, while clients are appropriately cautious about making large investments in this environment, they recognize AI's transformative potential and the value of strategic partners. This transformative potential is reforcing our industry's first principles, which underpin our evolving posture as an AI builder. The industry's first principles were born out of an enterprise reality.

Technology was so transformational and complex that companies needed help with optimizing the use of technology to meet their business objectives. IT services companies emerged to solve these problems at scale and over time, helped create many of the greatest business architectures over the last 50 years.

With AI, the fundamentals are shifting. Software is penetrating deeper into enterprises and our clients now expect more value and measurable outcomes. The old fundamentals are still relevant, but there must be reforge for a new reality. Cognizant has already embarked on this transition, which demands four significant shifts that redefine the role of IT services firms.

First, we are evolving towards owning the full stack of capabilities required to design holistic bespoke AI systems from a system integrator to an AI build.

Second, we are reimagining our talent moving away from the traditional pyramid towards interdisciplinary teams that operate at the intersection of domain operations and technology.

Third, we are shifting our economics from labor base to outcome-based models that align our success directly with our clients. Our combined fixed price and transaction-based portfolio has continued to grow in proportion over the past 3 years, reflecting our ongoing focus on driving nonlinear revenue opportunities.

And finally, we are evolving away from simply delivering projects to underwriting operational results for our clients at scale, taking full accountability for the business impact we create. Last quarter, I talked about the velocity gap the gap between massive AI infrastructure spend and the business value realization. And our Cognizant's mission is to be the AI builder who bridges this gap.

Our AI builder stack is the connectivity tissue that translates our strategy into measurable client outcomes. It combines our proprietary methodologies and the science of context engineering with a curated ecosystem of strategic partners and our own differentiated platforms and IP. Our vision is to reimagine enterprise operations, rebuild workflows and break functional silos to unlock AI native ways of working. We aim to do this by bringing human effort and Agentic capital together in a managed governed and a client contextual delivery model.

Some of our pioneering clients have started to progress from AI productivity to unlocking new experiences, products and services. Platforms are key to our AI builder stack. Fueling our platform strategies, our award-winning AI Labs, which was awarded 3 new patents, bringing its total number of patents to 65 in the U.S. and the 88 globally. Our AI lab continues to sense the future and partner closely with our clients, platforms and products group, and solutions teams to translate frontier research into industry relevant use cases.

To complement our internal investments, we launched the Cognizant innovation network, a new corporate investment arm that will back early-stage AI start-ups. We plan to initially focus on investments in AI, data, cybersecurity and cloud technologies and portfolio companies will gain direct access to Cognizant's deep industrial expertise and its enterprise client base, creating a powerful ecosystem for mutual growth. We are progressing towards the AI builder vision through our 3-vector strategy. AI-led productivity, industrializing AI and identifying the enterprise.

To date, we have well over 5,000 AI engagements across 3 vectors, up from approximately 4,000 exiting December. Beginning with Vector 1, we are addressing a multitrillion dollar opportunity of AI-led productivity across several value pools by helping clients, building classical software in new ways, accelerate software development, eliminate technical debt and modernizing legacy systems.

Our differentiated approach to autonomous software is rooted in engineering-led productivity powered by leading strategic partnerships like Entropic Claude, Google Gemini, Microsoft [indiscernible] and copilot Davin and open AI codecs. This approach has enabled nearly 40% of our code to be AI assisted. Cognizant platforms play a critical role in scaling these productivity [ grains ] by accelerating software development with Flow source, reverse engineering legacy code using agent-based capabilities through Sky grade and automating incident management with neuro IT operations.

A great example of our platform strategy at work is with one of the nation's largest health companies where we now underwrite the integrity of their claims process. Our AI solution automates the validation of over 54 million provider contract updates annually, directly reducing revenue leakage and solving a problem that was previously intractable at scale. Some of the early value pools in Vector 1 where we are seeing client momentum are related to legacy debt takeout, like mainframe modernization, SAPs for HANA migrations, autonomous software engineering, digital workplaces and autonomous infrastructure services. For example, we are working on a highly complex true blue field as for HANA transformation at a global scale, focused on modernizing the enterprise core for the North American global pharma leader.

What really sets this project apart is our use of a customized AI accelerator that automates both business and IT data validation, replacing a fragmented manual process with a scalable audit ready and a robust solution significantly cutting validation time and effort. For a leading European telecom operator, Cognizant delivered an AI-powered Oracle cloud ERP transformation, unifying finance, procurement and supply chain on a single cloud-native platform, achieving 25% faster time to market and 40% faster deployment through agentic AI and automation. And with Daimler Truck, we will use Cognizant WorkNEXT to transform and modernize its global workplace services.

Our multiyear partnership aims to leverage artificial intelligence and automation to enhance workplace operations across their global factories and offices. As our AI productivity capabilities mature, we are increasingly applying token metering at a project or an individual level to provide early insights into usage patents model management and optimization of infills costs. Vector 1 continues to be a primary driver of our large deal momentum. And as a result of the cost savings and shared productivity generated in Vector 1, we're starting to see increased velocity in Vector 2 and 3 opportunities.

Let me share some examples. In Vector 2, as we integrate enterprise AI into enterprise landscapes, platforms provide the foundation to move AI from proof-of-concept into production at enterprise scale, managing the full agent life cycle with neuro AI engineering and context engineering. This spans several areas, including data engineering, AI, foundry, cybersecurity and integrating AI into the infrastructure and cloud stacks. As an example, with data engineering for a leading U.S. health care client, we deployed an AI-based data validation system to optimize the distribution of pharmaceutical shipments. The solution uses predictive models to validate data before dispatch reducing downstream errors in the logistics chain and improving reliability across its operations.

One of the value pools we see in Vector 2 and a key element of our AI builder stack is context engineering. Cognizant's approach to contract engineering is to build native work graphs by going deeper into how humans work, make decisions and navigate exceptions in their daily business processes. We're also applying context engineering at a top wealth management firm with an advanced proof of concept where AI agents are being designed to work alongside financial advisers handling routine interactions and back office tasks so that financial advisers can focus on client-facing activities.

Finally, in Vector 3, we are accelerating development of our AI native products to unlock new agentic labor pools across vertical and functional domains and into core operations. The value pools in Vector 3 are significantly expansive opportunities across business operations of enterprises to embed Agentic capital for productivity, experiences and new services. In health care, for example, we are developing agentic solutions that accelerate and improve the accuracy of prior authorizations to support better patient outcomes.

Additionally, we are building on our TriZetto product portfolio in a strategic partnership with Palantir to advance an outcomes-based intelligence platform that embeds AI-driven decisioning directly into health care operations. We're also sensing a broad structural shift as AI moves beyond digital workflows to governing physical systems and environments and infrastructure. This is accelerating the convergence of physical AI agent AI and governed enterprise intelligence enabling autonomous operations across sectors.

Cognizant is investing in the architecture platforms, partner ecosystem and industrial domain expertise for physical AI. Business operations-led offerings are central to this evolution. We're expanding AI-enabled services across sales, finance, marketing, service operations, horizontally and health care financial services and banking operations on the vertical stack. Examples include the recent launch of autonomous customer engagement with Google to support outcome-based human AI workforce models across industries and the combined value proposition of TriZetto and Palantir to identify health care operations.

Across all 3 sectors as the importance of platforms grows, we are evolving our commercial models towards fixed and outcome-based pricing, enabling Cognizant to recognize the added value of assets, IP and accelerators that we bring. This is an important pillar of our first principles, shifting our economics to managed services and outcome-based models. Consistent with the shift, we delivered 2.5% and 5% increases in trailing 12-month revenue and adjusted operating margin per employee, respectively.

We are beginning to see the emergence of AI infused rate cards where pricing reflects a blended model of human effort and digital effort with several clients, we are exposing tokenized rate cards that prices work along a continuum from fully human-led discovery to hybrid to increasingly autonomous agenetic delivery. This model is intended to turn our outcome-based economics into a true partnership that aligns value creation with shared results. Execution across the sector requires the right organizational structure and a powerful innovation and talent. This brings me back to another important element of our first principles, reimagining talent away from the traditional pyramid and towards interdisciplinary AI-augmented teams. To fuel the shift, we have launched an integrated AI skilling stack for our entire organization. It begins with our AI Builder career program, which maps every role at Cognizant to a future-ready AI family, job family with defined pathways and targeted learning plans aligned to how [indiscernible] evolving. This is powered by Cognizant and SkillSpring, our new AI native learning platform designed to redefine learning in the AI era and cultivate AI-ready talent at scale for our associates and our clients and progress is being tracked for each associate's personal AI fluency dashboard.

A real-time context engineered view of AI readiness across various dimensions, including AI skills and proficiency, training and certification, AI tools and token uses innovation and project experience. To enable us to execute on these principles with the speed and the agility of the market demands, we are initiating a new program called Project LEAP. This program is designed to accelerate our transformation to the operating model of the future by funding investments in our AI capabilities and partnerships, integrated offerings and platforms, reshaping productivity and upskilling our workforce. By fostering a workforce that is AI-enabled and equipped with future-ready skills, we aim to create a more agile, scalable and cost-effective operating model.

Even as we make these changes, we are continuing to invest in growth through acquiring new talent. We hired around 20,000 freshers in 2025 and plan to hire a greater number in 2026, providing a strong pipeline of future talent aligned to how work is evolving and shaping a broader pyramid with a shorter path to expertise. The LEAP program reinforces our commitment to be in the winner circle of revenue growth and supports our journey of expanding margins.

To conclude, I want to leave you all with this. Our conviction in the long-term opportunity emerging with enterprise AI adoption has never been stronger. In our industry, the real work happens inside complex systems across legacy environments, regulated processes, global teams and mission-critical operations. Large enterprises do not transform overnight. The undeniably need trusted partners who understand their systems, context, risks and people. And that is the role we intend to play as an AI builder, bridging the gap to enterprise value. To win, we must move fast and stay agile, which is exactly why Project Leap is so critical.

We are reforging our first principles, enabling an AI Era future operating model, equipping our go-to-market teams across the 3 vectors. Adopting new engagement models to deliver value to clients and adopting talent through a blend of digital and human effort. We remain confident the portfolio and capabilities we are assembling can drive sustained progress towards Winner Circle performance including top-tier growth, consistent margin expansion and EPS growth outpacing revenue growth.

Before I turn the call over to Jatin, I want to thank our associates for their dedication, our clients for the continued trust and our shareholders for your confidence as we strengthen our foundation to create for durable long-term value.

Thank you, Ravi, and thank you all for joining us. As Ravi noted, our first quarter results demonstrate that our AI builder strategy is resonating in the market. In Q1, we delivered revenue growth in the upper half of our guidance range basis points of year-over-year adjusted margin expansion and adjusted diluted EPS growth of 14%.

First quarter bookings growth of 21% was 1 of our strongest in recent history. This performance demonstrates our focus on execution and our ability to deliver value for the clients. As Ravi mentioned, the market remains complex but the dynamics are not universal and vary by industry. Financial Services is benefiting from robust investment cycles, while policy changes are creating regulatory uncertainty in key areas of health sciences. In production resources, trade policy uncertainty and supply chain disruptions remain realities. That said, broadly speaking, we believe the shifts we are seeing in client demand play to our strengths. And we remain confident in our position as a strategic partner to our clients as they navigate a complex macro environment and the rapid pace of AI innovation.

Now moving on to the details of the quarter. In Q1, revenues of $5.4 billion grew 3.9% year-over-year in constant currency, driven by a ramp of large deals across our North America region and Financial Services segment, along with the strong performance in the U.K. We have seen increasing demand for our AI and analytics services. driven by AI readiness and innovation budgets. Growth also benefited from revenue from third-party products associated with our integrated offering strategy, and inorganic contribution from our 3 cloud acquisition.

By segment, Financial Services led with over 10% year-over-year growth in constant currency balanced across banking, financial services and insurance customers. We saw both healthy discretionary spending and sustained large deal momentum driven by North America. Health Sciences performance remained resilient. Growth was negatively impacted by approximately 300 basis points year-over-year due to a lower revenue from third-party products associated with our integrated offering strategy. Excluding this impact, services in health sciences grew at a similar level to the company.

Products and Resources was stable despite headwinds from macro geopolitical and trade policy uncertainty. We continue to see emerging client demand in areas such as predictive supply chains, agent commerce and hyper personalization. Use cases where AI has the opportunity to create real differentiation. Physical is an early stage but fast-moving category, and we are positioning ourselves to capture this opportunity as client adoptionaccelerates.

Within communications, media and technology, our revenue with technology customers continues to grow. AI adoption is driving demand for engineering, modernization and platform services. In the comms and media sector, the environment has been more measured with added pressure from client-specific dynamics tied to strategic shifts at a large customer. In Q1, segment growth was driven by revenue from third-party products associated with our integrated offering strategy, which contributed approximately 10 percentage points of growth.

Turning to bookings. We delivered another strong quarter of large field bookings. We signed 7 large deals, each with TCV of more than $100 million in Cuba, including 1 mega deal with TCV in excess of $500 million. On a trailing 12-month basis, bookings grew 11% and represented a book-to-bill of 1.4. Annual contract value was flat as deal duration increased in the quarter, reflecting large deal mix and continued softness in smaller discretionary projects. Our pipeline remains healthy and broad-based. We continue to see strong demand for cost takeout, vendor consolidation and AI-led services.

Moving on to margins. Q1 gross margin decreased by 80 basis points year-over-year, reflecting impact of our integrated offering strategy and increased compensation cost. We remain very focused on driving gross margin improvements over time. This is an important objective of the project lead program. First quarter adjusted operating margin of 15.6% increased by 10 basis points year-over-year.

Our ongoing focus on operational efficiency and benefits from the Indian rupee depreciation helped to more than offset the impact of our integrated offering strategy M&A investments and increased compensation costs. Now to additional details on EPS, cash flow and capital allocation.

First quarter adjusted EPS was $1.40, up 14% year-over-year. DSO of 84 days increased 3 days sequentially and year-over-year. First quarter free cash flow was approximately $200 million, impacted by a larger bonus payout this year and in line with our expectations and typical Q1 seasonality. During the quarter, we returned about $600 million of capital to shareholders through share repurchases and dividends. We ended the quarter with cash and short-term investments of $1.5 billion or net cash of $949 million.

Now turning to guidance. For the second quarter, we expect revenue to grow 3.2% to 4.7% year-over-year in constant currency. This includes approximately 150 basis points from our recently completed acquisitions, including a partial quarter contribution from Australia that we just announced. Our second quarter guidance includes a more cautious near-term view of discretionary spending based on recent global events and trends.

Our full year revenue guidance is unchanged at 4% to 6.5% in constant currency. The macroeconomic environment remains dynamic, and our guidance reflects a range of outcomes. We expect large steel ramps and 2 full quarters of Astra contribution to be meaningful second half drivers. At the midpoint, we assume some improvement in discretionary spending in the second half of the year compared to our Q2 assumptions.

Our strong bookings momentum, along with 1.4 book-to-bill ratio give us confidence that we are winning in the market. Our full year guidance assumes recently completed acquisitions will contribute approximately 150 basis points to revenue growth, reflecting contribution from both CreeCloud and Austria. Beyond this, our M&A pipeline remains healthy and active, and we see a number of interesting opportunities that are consistent with our AI builder strategy.

As always, we'll be disciplined and deliberate but remain well positioned to act if the right opportunities emerge. Now a few more details on Project Leap. This is an important initiative to accelerate our path to a more agile and AI-enabled operating model of the future and improving our cost of delivery. The program is expected to deliver savings in 2026 of approximately $200 million to $300 million with a full year benefit in 2027. We anticipate approximately 2/3 of the savings generated by Project LEAP will be directly reinvested to support future growth across integrated offerings, AI capabilities and partnership and roughly 1/3 toward upscaling our workforce, all while maintaining an active and strategic M&A posture.

The expected savings generated from the program, net of investments are enabling us to raise our 2026 adjusted operating margin guidance range to 16% to 16.2% and which represents 20 to 40 basis points of year-over-year expansion. This is on the top of 50 basis points of margin expansion we delivered in 2025 and in line with our long-term aspiration to expand margins. As part of this program, we expect to record costs of $230 million to $320 million, which substantially all incurred in 2026. This consists of $200 million to $270 million of employee severance and other personnel-related costs and $30 million to $50 million of other charges. This cost will be adjusted in our non-GAAP financial measures.

As Ravi noted, we will hire more recent college graduates this year than last year. Our free cash flow conversion guidance for the full year remains 90% to 100% of net income. Tax rate guidance is unchanged at 25% to 26%. In our expected weighted average dilutive share count is approximately $473 million, down slightly from our prior estimate due to the pace of repurchases in Q1. This leads to EPS guidance of $5.63 to $5.77, representing 7% to 9% growth.

For 2026, we still expect to return approximately $1.6 billion of capital to shareholders, including $1 billion towards share repurchases and the remainder towards our regular dividend.Finally, we continue to make progress and advance on our evaluation of potential primary offering and secondary listing in India. We remain committed to acting in the best interest of our shareholders and will provide updates as appropriate.

To close, we are delivering on our commitment to stay in the winner circle. In Q1, we grew revenue at the top of our large cap peer set, posted our strongest booking growth in recent history, expanded adjusted operating margins and delivered double-digit earnings per share growth. While the macro environment remains uncertain, our momentum is clear, and we believe we are winning in the marketplace.

With that, we'll open the call for your questions.

[Operator Instructions] Our first question today is coming from Jason Kupferberg from Wells Fargo.

So bookings, a clear highlight this quarter. Wonder to see if you had any color on how much of the bookings were new versus renewal, anything on ACV growth how that looked in the quarter? And just given the fact that there has been a little bit of softening on the discretionary side, it sounds like in certain verticals.

I wanted to confirm, Jatin, if I heard you correctly, that the midpoint of the '26 guide now assumes a little bit of improvement in discretionary spending in the second half. Maybe you could just elaborate on that a little bit? And then I have a follow-up.

Yes. So Jim, we don't exactly break out new versus renewal. But this year -- I mean, for the quarter, it's been very healthy. And I would say the growth, especially of the large deals is driven by the newer opportunity in either existing customers or the new customers.

In fact, just to add to Jatin, this is the second quarter in a row, we've had robust bookings. The new proportion is as healthy as it was in the past. In fact, the top 7 deals, which are more than $100 million, 1 mega deals, more than $500 million a 70% increase in TCV on the large deals. I think it's been a good quarter for bookings, 2 quarters in a row. This is probably the highest bookings growth we have seen since I've been on board 3 years ago, since 3 years.

Okay. Okay. And there was some commentary from one of your large competitors last week talking about AI resulting in increased competition, more pass-through of productivity gains to clients. I mean, you guys have been talking about that pass-through and the AI assisted coding for a long time. But are you seeing competitors broadly engage in any additional level of contract pricing that you might characterize as a rational?

Yes. [indiscernible] the way I see it is, unlike in the past, where pricing was determined by the unit price, which is billings equivalent. The race now is about the number of units and how well we could deliver with lower number of units for the same output for the same outcome. And that is based on how much productivity you can derive out of AI usage in your software development cycle.

So we feel very confident because 40% of our software development cycle is assisted by AI. We have infused AI into our rate cards. So when we are up for a consolidation opportunity, we seem to be in the winner part because we are able to share the productivity and also keep it for ourselves. In fact, a bit to date is actually very healthy over the last 3 years, the means we have been working on.

So we seem to have got this work rhythm of autonomous software engineering, as we call it, and we seem to be doing well. So I wouldn't -- I mean there is productivity sharing with clients, but you're going to see that as an opportunity to win more and you will see that as an opportunity to create more momentum for ourselves. So that's how we are seeing it. Now that's on the old stuff.

On the new stuff, there is in Vector 2 and vector 3, as I call it. It is new work. We also see unlock of legacy modernization, which is not consolidation. It is actually net new business which kind of got locked because customers were not willing to pay that much to modernize the legacy. Now they're actually throwing that in the mix. And that is actually a business case of how much they spend to how much they could potentially spend using AI to modernize it. So while there is price pressure on it, I would say it's an opportunity not to look at labor costs. It's an opportunity to look at -- how much of the number of units you could optimize using AI.

Yes. And Jason, to your earlier question on the visibility of second half and how do we see our guidance range. Let me break it down. So definitely, the environment around us, the macro has significantly more uncertain than what it was in the beginning of February. And therefore, I mentioned in my opening remarks that there are a range of outcomes which are possible across the guidance range for the full year. What gives us confidence for the second half are essentially 2 things: the large deal wins that we have had in quarter 4 and in quarter 1, which continue to ramp up and will reach its full potential -- their full potential in starting June, July. And therefore, that's one lever. The second is acquisition like Estia will come full on stream from quarter 3 standpoint, it would be a partial revenue in quarter 2.

So these are 2 additional sort of drivers for a stronger second half than assumption than what it is. As I mentioned, the midpoint does assume a little better discretionary environment than what we assume for quarter 2, which is sort of impacted by the current environment, but we remain confident as we walk through the rest of the year. And finally, we have had very strong bookings for quarter 1, which means we are mining in the marketplace even as in the uncertain environment, customers are choosing Cognizant as a partner of first preference. And that is what helping us continue to lead in this environment with the sense of Velocity and confidence.

Also a lot of large deal transitions, which are happening now between quarter 4 and quarter 1 will start to unlock in quarter 2 and quarter 3. So this is literally production capacity already in there, we are not making the money we are incurring the cost. Now as the transitions get over, we start to accrue the dollars. So that's actually another tailwind to our journey in the second half.

Our next question today is coming from Jim Schneider from Goldman Sachs.

I was wondering if you could maybe kind of unpack the comments you made earlier, Ravi, around the token usage that you're seeing in terms of token metering and also some of the productivity or benefits you're starting to deliver. Just wanted to clarify 2 things. One is, are you seeing with the increased sort of productivity on units delivered to your customers. I would have thought that you would be seeing if you're keeping some of that benefit for yourself, a little bit better margin leverage as a result of that as you're getting some revenue growth? Or is that being masked by the start-up cost on longer-dated outsourcing deals you just kind of talked about in response to your last question.

And then separately, I was wondering if you could maybe address how you expect sort of the token usage to sort of play out in terms of how you build customers? You talked about AI type rate cards -- but are token costs being directly billed through to clients today in terms of time and materials contracts.

Great question. Great question, both of them. Now let me first get to the second one. Token metering is a reality, both at a project level and at an individual level. We have token metering for fixed price programs as well as for time and material. For fixed price programs, we have the opportunity to reduce the cost and keep the margin with us. For new deals we do this kind of links back to the first question, we actually have the opportunity to outpace the productivity we give to our clients and therefore, keep it with us. A bit to date over the last 3 years is very healthy. So I'm very excited about the fact that has leverage for margin accrual in the future.

Of course, it has start setup costs, which we have to establish at the start. So there is an upfront cost attached to it, but there is downstream savings. On time and material, tokenized rate cards. We are starting to see a pattern. I'll give you 1 example.

One of the rate cards. I am establishing, which is a template we're taking it to the street is A0 is completely human effort. A1 is effort, which is done by humans verified by AI. A2 is effort effort delivered by AI, verified by humans. A3 is autonomous digital labor. Now when clients do this, you could meet the capacity they have bought from a frontier motor company like Anthropic or open AI directly. In which case, we are responsible for the human effort and the clients are responsible for the digital effort. But clients have started to see that they are not able to optimize the digital efforts.

So some clients are coming back and saying, you know what, why don't you take care of the human and the digital effort. You open the tap on compute, you open the tap on Ms. and you deliver the service, and we don't want to manage the economics. Already, we're talking about AI ops, AI FinOps. You manage the economics and digital labor or human level, it doesn't matter. In fact, 1 of our -- 1 of our research papers talks about a cognizant cognizant unit of measure, which we call it as -- it's equivalent to the function point measure, which is equivalent to what we do in digital labor. So effectively, this is evolving. We are ahead on the curve both to take the accountability of digital and human labor for ourselves or if clients want to take the accountability for themselves. So time and material comes in 2 forms. I could take care of digital labor and human labor. And I could take the sizing and the economics of digital labor and create throughput for our clients. This is something evolving and some clients are already proposing this. And on fixed price deals, of course, we want to share that productivity with our clients.

So that's how we are seeing this. And it's not far off when we're going to see rate count for this rate card, which is digital and human label put together more mainstream. As we go forward. And that gives us an opportunity to actually deliver both human and digital labor through the books of Cognizant. We already have arrangements with the fronter model companies to do that. One is to take care of our developer community, which uses it, but also for client work, which we can deliver.

Jim, I'll quickly cover the question on gross margin. Essentially, we -- in Q1 was an investment mode a little bit on gross margin across 3 different dimensions. The first was surely, the investment in the bench. And if you see, we have grown sequentially in headcount, and we have grown year-over-year in headcount. And as we have continued to hire the fresh college radiates into the mix. We have invested a little bit of utilization. So that's 1 reason why gross margin is lower.

The second, we spoke about this integrated offering. We -- every time when the industry sees a new element of service delivery coming to customers, they expect service provider like cognition to act as a system integrator. And to that extent, you see that you have a higher element cost as part of this integrated service offering that you have, and that has been slightly higher in quarter 1. And -- but that's an investment because you almost always see a significant follow-through revenue coming through services when you anchor yourself through that early offering. And third is the salary increase that we gave on first of November, so there is a 1-month impact sitting there. So a combination of these 3 factors have led to a slightly lower gross margin in -- in our earnings call as well as through the quarter, we spoke about the volatility that would probably see as a result of this investment in quarter 1, but we are confident that the number will continue to improve through the course of the rest of the year. the project ambition is to really drive significant cost savings through cost of delivery model. And that should also help the gross margin as we execute for the rest of the year.

Next question today is coming from James Faucette from Morgan Stanley.

Appreciate all the color and detail today on current conditions, et cetera. I'm wondering if you can talk a little bit about what you're seeing in terms of valuation and how you're thinking about the price of acquisitions that you're looking at, particularly as you seem to be looking to add incremental capabilities to the Cognizant base? And I'm just thinking about how we should think about your commitment to spend, what portion of free cash flow and the impact on the inorganic contribution on a go-forward basis?

I mean I'm going to ask Jatin to add. This is a phenomenal time to create value from M&A in line with our reforged first principles, which is about having a platform player, managing business on outcomes versus effort and AI enabling our offerings. So if you put all of that together, we have some exceptional opportunities in the market. We also have the ability to anchor this on new pillars in the mix, which will give us expansive opportunities.

So we are we are not doing this in a tactical way. We are doing this in a very strategic way of filling the boxes for being an AI builder. That's what our -- this is. Just to give you a sense, today, the 1 we announced does data data center build-out services, workplace services, AI infrastructure build out services and network services. So it's Atria is a phenomenal opportunity to anchor a complementary piece of work, which is which is attached to the infrastructure services where Cognizant is delivering very well. So where will we anchor this. We'll anchor this on platforms, on outcome-based models. In fact, Atria delivers work on per user basis, not on effort. I mean they do workplace services. They also do data center build-outs in an outcome-based model.

So we think that's a unique opportunity. If there are platforms in the market, we're going to evaluate and look for it because the platform play will allow us to go through -- go to the outcome model, transaction-based pricing and outcome-based pricing. In and around set -- in and around TriZetto, we see a ton of opportunity. I mean, our TriZetto business now is growing much, much faster than what I saw in 2023 when I came on board. And it is highly profitable. And health care has a strong moat, a defensible mode. So we are actually we're actually looking for layering it around.

In fact, one of the reasons why we partnered with Palantir is to create the opportunity to drive the health care payer control points for medical loss ratio performance and payment integrity and real-time cost intelligence and network performance and all of that. So we have specific areas where we think we want to do M&A, which will substantiate our endeavor to be a platform company and an outcomes-based company in the AI era. And we also believe it will uniquely give us an opportunity to create durability of our earnings.

So that's how we are seeing it, and this is a good time for a value player. So we are continuing to -- we have a very healthy pipeline. We'll continue to evaluate value assets, which are available in each of these pillars I just spoke about.

And just from a capital allocation standpoint, you know we generated $2.5 billion of free cash flow. Last year, we returned close to $2 billion to shareholders and roughly $500 million, $600 million or $700 was invested into 3 clouds, which technically closed beginning of this year, but was announced in 2025 -- this year, again, $2.5 billion, we have committed $1.6 billion to be returned to the shareholders $1 billion by share buyback and $600 million or in dividends. Of which we have now used about $600 million from the remaining $1 billion for Asia. And we have, therefore, sort of fuel in the tank, and we have a very healthy balance sheet to leverage in a very attractive opportunity [indiscernible]

That's great. And I just asked that you guys have been really front-footed, both in terms of like your own development. prioritizations and how you're trying to implement AI. And then I think your commentary just now on how you're looking at acquisitions and some of the benefits that they provide further bolsters that view. What types of customers are you seeing either generally or what kind of characteristics do they have that are willing to engage with you and kind of match your march forward right now? Where are you seeing the best traction? And where should we look for examples of success patterns?

Yes. I would say -- I'll just highlight quick themes. Financial services is at double-digit growth, very, very excited about it. Not only are they doing Vector 1, they're innovating new products, new services. Vector 1 is more productivity led and are willing to experiment with us. In fact, 1 of the things I mentioned in my earnings script is about opportunity with management company to apply a genetic on their wealth advisers so that they could deliver more innovative products.

So financial services is right up there. The second I would say is consolidation opportunities. Consolidation opportunities, I mean, every customer, every Fortune 500 Global 2000 company has a huge set of providers. This is a -- I mean, they have accumulated they've created a big network of providers over the last 25, 30 years. This is their opportunity to consolidate and get some productivity benefits. We are on the front of it, and we are winning a lot of it.

So that is the second. The third I would say is unlock of -- the third value pool, I would say, is unlock technology debt. I have great momentum on mainframe modernization. Just to give you a sense, 1 line of mainframe code used to cost $10 to refactor to new age, say, new edge software it now costs $150. So we have a unique opportunity to unlock. And this opportunity didn't exist before because fiber knowledge was missing, there was cost, financial cost to modernize, and they were legacy skills were missing. Now all of that is out of the window.

So if you unlock that, that is trillions of dollars. So we're seeing that as a good team. Then there is operations-led AI. I mean that's why my DPO business is close to double-digit growth because operations of enterprises are going to be embedded with this new software, which operations didn't have a chance -- and if I have to pick specific areas, customer service is the top area where we are seeing this. Employee Services is the second area we are seeing and then traditional areas like financial systems and legal systems. These are places where we historically didn't see a lot of classical software embedded. So we're starting to see that.

One of the things I highlighted in my earnings script is physical AI. I mean it's a leap frog for traditional industries with physical manifestation to actually invest into digital enhancement of physical objects and we're seeing quite a bit of that. So I think we are at that inflection point now. to take productivity and create elasticity of consumption of software, classical software and use new software, which is written around the neural networks, to an expansive opportunity in enterprises and integrate the 2 and reinvent and reimagine businesses. I mean this is -- this is a fabulous opportunity for system integrators to be those builders. And I'm actually more optimistic than I was before on the opportunity in front of us.

Our next question is coming from Tien-Tsin Huang from JPMorgan.

Great. Just want to ask on Project LEAP, if that's okay, just the offensive or defensive nature of it? What prompted it, the scope of it and what outcomes we can expect in the short and midterm from Project LEAP.

I'll kick off, and I'll get Jatin to add. I mean, we have a mental -- a frame of what our future operating model looks like, and I'm pretty confident that operating model is we are on the journey to get to the operating model, LEAP is to make sure that we get there fast. It's our opportunity to resize our pyramid with a broader parameter. That's why we're hiring more school graduates, more early careers. And short on the height of the pyramid so that you get to expertise much faster. That's our model. It is margin accretive because the more you brought on the pyramid the more you could deliver the services in a more AI native way, if I may.

The second important thing is this allows us to invest into the platforms AI enabling the enterprise and tokenizing the enterprise, as I call it. It allows us to do that. So we have measured the savings. It is in the range of $200 million to $300 million this year, and this is partial because we are already in the middle of the year, and we have couple of months to complete this process and probably 3 to 4 months of impact.

So it has a much higher impact next year in 2027. So not only are we rightsizing the pyramid. And remember, we're also seeing our future offerings are not effort-based, they're outcome-based, more and more. I mean, we'll see a mix of it in the transition. So when we get to that new operating model quicker, we are going to seize these opportunities faster and we'll be we'll be having a more optimized operating model. And we will have a kitty for investing into our future. So that we can seize the opportunities ahead of others. We also said in our Investor Day that we will have an expansive margin trajectory, which is what we intend to. I mean last year, 2025, we did 50 basis points in spite of the M&A and the investments -- this year, we have upped our margin guidance to 20 to 40 basis points increase from 10 to 30.

So we are constantly on that trajectory to keep increasing our margins keep delivering productivity to our clients and be in the winner circle of growth. So this allows us to do all of this in a quicker, faster way. We get to that future operating model, which we have in [indiscernible]

Our final question today is coming from Surinder Thind from Jefferies.

Ravi, can you expand upon the last point of -- what is the benefit of showing margin improvement in the current environment relative to your ability to invest. Why not just maximize every dollar of spend maybe broaden the spend across what I would call more of a VC type strategy where you take more bets because the pace of change is accelerating -- and so as you try to build and adjust the model like.

I think you're spot on. If you look at it, we're going to save $200 million to $300 million this year, which is just a few months. And you know in 2027, we have a much bigger opportunity. But we're investing back the rest of the money to generate growth opportunities and be agile enough in the market to generate more growth opportunities. So you're exactly right. So we're investing more into growth and we are contributing some into our expansive margins. So the idea of doing this is growth and be in the winner circle.

Thank you. We've reached end of our question-and-answer session. I'd like to turn the floor back over to management for any further closing comments.

Thank you so much for listening to us. I mean we're very excited about our quarter 1 performance. Very excited about the bookings momentum we've had and the tailwind we have for the rest of the year and, of course, are anchored to the AI opportunity and getting their fast with the LEAP program and keeping our thesis of being in the winner circle from growth, expansive margins and EPS growth being higher than revenue growth.

That's our endeavor, and this will allow us to create sustainable durable earnings for the future.

Thank you. That does conclude today's teleconference. You may disconnect your lines at this time, and have a wonderful day. We thank you for your participation today.

Thanks for joining us this afternoon. My name is James Faucette, Senior IT services analyst here at Morgan Stanley. And I'm very pleased to have John Dalal, CFO of Cognizant that will be chatting with us this afternoon. Before we get started, I do have a quick disclosure to read. Please see the Morgan Stanley research disclosure website at morganstanley.com/researchdisclosures. If you have any questions, please reach out to your Morgan Stanley sales rep. So thanks for being here today. I really appreciate you guys making the trip. Always excited to hear kind of what's happening with Cognizant. Certainly, there's been a lot of work underway for a few years with the management team very focused on AI development, et cetera.

But -- maybe to start, why don't you give us a little bit of a preamble, talk about how you put together your outlook for calendar year '26 and how you're thinking about organic growth generally?

Sure. Jim, thank you very much for having us. Wonderful to be here with meeting all the investors. Let me start with our -- the whole assumption around guidance range. And as you know, we guided 4% to 6.5% total growth for 2026, of which inorganic contribution is roughly 150 basis points, of which 100 is done and 50 to go. Now at the midpoint of our guidance range, we assume that the environment continues to be what it was at the time of giving the guidance. At the upper end, we expected the environment to improve during the course of 2026. And clearly, at the bottom end, we have assumed that the environment falls off from where we started, and that could be the lower range. So that's the overall thinking as we think about 2026.

So with that 1%, 1.5% of organic growth, it seems all right, maybe slightly better than what we've heard across the market as a whole. How much do you attribute that to market share gains versus a broader improvement in client spending? And let's start there, I guess.

Yes. So if you -- the midpoint of the organic range, so if you see the total range is 4% to 6.5% midpoint is 5.25% and therefore, the organic is 3.75% for 2026. It is a combination of both. We have seen some amount of gains coming from discretionary spend, which is in areas like BFSI. The market is expanding, and we have participated better than others in that discretionary spend. And through our large deal wins, there is, of course, market share gain as we have consolidated our position in some of our key accounts. So that has also contributed. So it's a little bit of both that we have considered for our midpoint of our organic guidance.

Got it. Got it. And -- let's talk about the competitive environment a little bit. What do you think is central to your ability to take at least some share in the current environment?

A few things. The whole cycle of winning more business in consolidation when the industry is going through consolidation is about your ability to execute well. We continue to win large deals, but more importantly, we execute them very well. So we won 27, 28 deals in '24 and similar number in '25. And we continue to execute well and customer then builds a sort of a pool of references come forward and therefore, you're able to win the next deal and deal after, et cetera. So the beauty or virtuous cycle of large deal wins is your ability to execute well. The second is you should be seen as a company of future and a company which is leaning forward in investments like AI to be seen as a partner of future. And very happy to share that I think we have built that image over the last 3 years that our customers perceive us as a company which is investing in future and that's the right partner for them.

Right. Got it. So let's talk about your clients and what they're communicating around their calendar year '26 IT budgets relative to particularly last year. Can you give a general assessment now that we're really past February, how budget growth compares this year, looks like it will compare this year to last year? And where are you seeing either by vertical or geography spending increase versus where would you characterize clients as still being more cautious?

Yes. I would say the environment remains what it was. So it has certainly not shown an acceleration from where it was. Of course, there are more macro issues that we have seen in last week that have surfaced so to -- as a short-term headwind, so to say. But the environment remains what it is. From an industry standpoint, we definitely see, BFSI continue to grow very well. For 2025, we grew 7%, and it exited at a 9% growth run rate. Health care, which is another large segment for us, has grown 5%, 6% in 2025. So despite regulatory uncertainty, that industry has done well. Where we do see a certain lack of discretionary momentum is products and resources and communication and media. In both places, we see more of a wait and watch or a slow growth scenario.

So can I delve in a little bit? So when you talk about like BFSI, like doing pretty well, that makes sense given kind of the way those end markets have been functioning. Health care always seems like it's a work in process there. But materials and that kind of thing, -- it seems like -- and maybe this is just an artifact of the stock market, but it seems like that market is maybe a little bit better. What do you think needs to happen there for that industry to start to feel a little more optimistic about their spend?

I think that industry, especially retail and manufacturing has been far more focused on the supply chain during the course of 2025. And with some stability emerging there, I think the priority will now shift to what they need to get accomplished on IT and new trends like AI for their businesses. So I think '25 was more about catching up on what they need to accomplish on the supply side.

And obviously, that was very disrupted or a lot of uncertainty introduced because of tariff and potential changes in tariff regimes, et cetera. So that makes sense. So let's pivot to everybody's top topic, AI and its implications. You've noted that gen AI-driven productivity is really being embedded into new deal economics. How are recent deal vintages tracking versus your internal efficiency assumptions? Are you performing as you expected when you put these deals together?

Yes, absolutely. And so this is something that we are very careful when we first review the deal before we submit our bids. We look at very carefully how the deal is constructed and what is the solution? Because you don't win some of these deals on price point. I mean that's one of the factors, but it's also how well need and how well thought through your solution is for customer makes a big difference. So one is we review it very well. And then we track it versus -- through a process called bid versus bid on a monthly basis, the CEO of the company and I review it every month. And we have done quite well as a portfolio. We are within a percentage range of the revenue that we expected from the deal. We are within percentage points within the margin that we expected. So as a portfolio, we continue to deliver well. We have a delivery excellence team, which continues to review the deployment of tools and productivity measures that we had anticipated in the deal and how they're getting deployed. So we have the orchestration behind to make sure we deliver to the promise that we made to the customer.

Got it. And how do you see that -- those types of engagements evolving? I mean, do you see customers be more willing to accept that? Or do you feel like they're being more demanding in assuming productivity improvements? And how are you navigating that? It's a very collaborative process where you work with customer. I mean, there is no deal which gets signed on the RFP being out and bids being submitted. There are at least 4 or 5 very detailed workshops that we do with customers to decide what is the right answer from a solutioning standpoint to very specific customer context. And therefore, there is very little surprise from a customer standpoint on what to expect when we start deploying the solution. There is over, I would say, last 5 quarters, there is a significantly higher openness and keenness to deploy the newer technologies and therefore, expect a greater productivity and total cost of ownership significantly better than what it could have been, let's say, in 2023. So there's definitely more keenness to deploy. Got it. So once again, sticking a little bit on contract structure. Fixed price revenue grew strongly in 2025. Looking ahead, where do you see the primary risk in these structures, whether it's in scoping, productivity assumptions, execution, change order dynamics. I think it's interesting to me at least that fixed price as something to do or not to do kind of ebbs and flows over the history of IT services. So where are we now in terms of where it could go? And what are your key risks?

I think the risks of -- and of course, rewards of fixed price projects remain, as we all know. But I think this is a better model for our customer when there are a lot of things changing around them, and it is difficult for them to anticipate the sort of total cost of ownership they would be able to achieve in a time and material basis. This is a certainty of outcome that we are able to give. Most of the time, we deliver to the point. So it's not that fixed price projects are significantly more profitable for us or significantly more loss-making for us. It is the certainty of outcome that is the best sort of offer coming through fixed price projects. And a changing technology landscape, we also like fixed price project because sometimes you assume that you will use a technique A and tool A. And then when you actually start executing, you get a better answer to the same problem and you can deploy it and get to the superior outcome. So the flexibility it leaves with partner as well as certainty it offers to the customer is something that creates the win-win for the fixed price project.

So do you see that continuing to grow as a proportion of your mix? And where can it get to, do you think?

Yes. We continue to believe that -- we believe that it will continue to grow. And over the last 3 years, it has moved about 6, 7 percentage, 41% to 47%, 48%. And I think a 2% to 3% increase every year is quite feasible.

And how important is moving to fixed price to achieving kind of your margin targets? We'll come to those more specifically in a minute. But is it a meaningful contributor or unlikely, hard to say?

It is definitely a good enabling block for us to have sort of a free hand to execute once we have signed up a contract. So it's a good enabler, but it also comes with a greater risk. So you could have, of course, a more cost deployment to get to the same outcome. So I wouldn't say that it's a nirvana for margin, but it is a good enabler or a good building block to get to a good outcome. We also make a good profitability on time and material contracts. So that's also fine. But I think the -- so I don't think we should drive fixed price projects just from profitability angle. It should be just the win-win that we -- it should be more about the win-win that we spoke about before.

So I'm hopeful that you can help address a little bit of a mystery or conundrum around BPO. BPO has grown roughly at 9% to 10% for the last several years, quite a bit faster, obviously, than the corporate average and certainly ahead of the industry. On the other hand, most investors really have concerns about gen AI disruption, particularly to this segment of the business. What has been driving that growth in practice? Is it net new demand, share gains? Are you being able to increase pricing, expanding scope beyond typical BPO? Just help us understand what's been driving the growth in BPO and how sustainable that might be?

Sure. So BPO has been a great adopter of gen AI technology. And while the initial perception was that because of gen AI adoption, the throughput that BPO generates would shrink or the volume that it operates will shrink. But effectively, it has been otherwise, meaning we are able to now generate more throughput with the combination of human and virtual agents, and that has been lapped up by customer. On top of that, we are also going out of -- after the total addressable market, which was not accessible to third-party service providers. For example, there is 100 people customer service work. Essentially, it would not have gone to BPO because it would have been too small.

Too small.

But now multiple thousands of such hundred people pieces of work are spread across many organizations, which are now far more amenable for an agentic deployment. You don't even have to worry about offshoring, et cetera. You could, but you need not, you could even deploy an agentic solution wherever that work is, and you could get significant productivity, like 50%, 60% productivity coming through that. And therefore, UC BPO has not only overcome the compression that was perceived as a risk to the sector to the revenue line, but is actually growing healthy as an industry, not just Cognizant, but other companies also in the industry.

Yes, for sure, for sure. So A similar question to fixed price. So in some of the other companies that are publicly traded and have financials in the BPO space, their margins tend to be pretty good, sometimes 400, 500 basis points above kind of even where Cognizant is operating. Is BPO for you being faster growth? Is it a margin contributor right now? How should we think about that?

Yes, absolutely. It's a great question. And BPO business is not only faster growing, but it's also better profitability for us. So all in all, it's a good portfolio contribution that if they go, grow faster.

That's interesting. Now on to the most the one topic that people tend to be most sensitive to, and that's discretionary spend, right? Because I think Cognizant over -- particularly the last 3 or 4 years has done a really great job, not only on BPO, we've just talked about, but also in some of the other work that you're doing on long-duration contracts and outsourcing and the like. But discretionary spend seems to be the volatile piece and can make the difference or at least make a big difference 1 year to the next in terms of growth rates or even versus estimates. It seems like there's been a little bit of spend improvement that has been visible in financial services. What types of projects in financial services are seeing the strongest demand? And what do you -- how do you view the durability of these key drivers?

So that's an excellent question. I think discretionary spend is what goes and sits on the layer of large deals that you win. So discretionary spend tends to give you a bump for the growth in a particular year. And as a historical reference, the discretionary spend really grew post-COVID and then by end of '22 or early '23, that started moderating. We definitely see the BFSI discretionary spend. As I mentioned, we grew 8%, 9% for quarter 4 of 2025. These contracts are in 2 or 3 buckets, but the one is really compliance-led work where a lot of change around BFSI continues to happen. And at some point in time, they need to catch up on that. The second is BFSI is an early adopter of technology. So they are now deploying AI tools and techniques in their workflows, in their functions, their customer service, their engagement with their end customers and so on and so forth. So that's second. And third is they -- in terms of some of the work that was accelerated on digitization post-COVID and put on back burner now is coming back in terms of how do I improve my customer journey for my digital bank, et cetera. So it's some compliance. Second is improving overall enterprise or functional strength of the BFSI organization. And third is customer journey. So these are the 3 things that we see.

Got it. Got it. I've been monopolizing questions for the last 20 minutes or so. Any questions from the audience, just raise your hand and we'll give you a microphone. So let's go -- just stay on the discretionary spend topic. So a little bit of improvement in financial services. Where else have you been seeing recently some improvement in discretionary spend outlook?

We have seen some improvement definitely on health on pharmaceutical side. There is some discretionary improvement. We have seen a little bit improvement also on consumer product side, which was actually clubbed with retail and manufacturing until, I would say, quarter 3 of last year, but it tends -- it is sort of coming back in terms of discretionary spend.

Got it. Got it. And any place that's getting worse on discretionary spend? I mean, I don't know how it could have, but...

I think I would say communication and media has been sort of volatile. So that's what I would say. I mean, one quarter is good, 1 quarter is a little uncertain.

Got it. Got it. And what do you think is driving the volatility there? Any speculation?

I think there are more company-specific, especially on communications side, there are company-specific initiatives that are taken to either significantly reduce cost or transform the organization. So I think there are various company priorities which are taking over the immediate IT spend decisions.

How much on uplift in discretionary spend? I think one of the things that investors have been keenly trying to watch is as companies and organizations move from AI pilots and trying to figure out what makes sense to move into production, et cetera, turning to the likes of Cognizant to help them with those development work, implementation, maybe even restructuring and rearchitecting some of the underlying systems and data, like what's happening from that perspective? Is that a meaningful part of your work or even of the discretionary business that's starting to come in the front door?

We definitely see that we are moving away from just POCs and let's create something which looks good but it's not scalable. It's small application. To real impact on business, we spoke about deployment of Agentic solutions to one of our large logistics customer in U.S. Very recently, we have rolled out a large Agentic solution for one of our customers in food processing industry. So there is now more meaningful deployment of Agentic solutions that we see. And we see that there is a unique role that companies like Cognizant has to play in terms of helping customers determine what is the right answer for customers.

Got it. I think we have a question in the back. to AI pipeline and production and just...

Status of gen AI is moving really quickly right now, right? So the velocity is very fast. Is that an impediment for -- how much of an impediment is that for your customers in terms of like waiting until kind of maybe the rate of improvement slows down before going through the long process of asking Cognizant to do something really big for the company and maybe an analogy is looking at past cycles where at the initial phase of the technology, that you have to wait for the rate of improvement to slow down somewhat before -- the bulk of your clients start to sign up...

I would say it is not so much about rate of improvement to stabilize before the deployment begins. I think it's bringing AI to B2B applications, which is just beginning to take off. So far, it was lapped up by consumers, but now we are beginning to see where the application is becoming real for B2B -- for large customers. So we have examples on a faster deployment of a package implementation. We have -- we have now examples or quick solutions regarding modernization of legacy code. And some of that is now beginning to getting picked up much faster than what it has been. So until 6 or 8 months back, the technology was for the technology's sake. Now it is really the practical application for a Fortune 2000 customers, which is coming to force. So we are very early in that cycle. And yes, there are a set of customers who are waiting to see what is the final outcome and then I will sort of go after it. But more than that, I think it's really making practical applications of AI for large-scale IT problems or IT challenges of a Fortune 2000 customer is just beginning to happen. So hopefully, it will take off in -- as we go forward.

And then a follow-up question. we just give him a -- just give you a follow-up question, and then I'll repeat it. -- so the question was how do these AI-related project sizes compared to traditional contracts?

Yes. They're still smaller in size, maybe $8 million to $10 million, $12 million in TCV, whereas, of course, in traditional size, you could have $200 million, $300 million, $500 million of contract. But that is not unanticipated for a new service. Almost every new service that we have ever sold as an industry has started on small scale. So this is actually the right size to anticipate as we look at a new technology deployment.

Got it. Question over here in the back.

Yes. So what is -- in your contracts to date or as you work with Agentic solutions in enterprises, what does the shape of that adoption typically look like? Do people start by building their own solutions? Are they using OpenAI's frontier? Are they taking out-of-the-box solutions from SaaS companies? Like where is it that enterprises are investing today when you see...

So it is different for different customers, as you would imagine at this point of evolution. But effectively, what customer starts by is that I have a business problem that can be solved by AI. Let me first think about what is the solution that I'm going to build. So you start with what is the compute that you are going to get, what is the LLM that you're going to deploy? Do you need entire LLM or you can build an SLM and filter it 20%, 30% to LLM and risk we can solve at the end to reduce the cost. Do I need agents which are branded agents from SaaS players or I could use a generic agent to be deployed and I could do from a cost standpoint, a combination of 30%, 70% or something like that. Then you go back and check saying, this is the data requirement, and this is a training requirement. How do I sequence my load on GPU from training versus actual usage. Then I start deploying. And then I think about how do I monitor the performance of agents. So this is the whole cycle. It is not going to one partner and asking the answer, but it's really working with the ecosystem and developing this whole integrated answer before the deployment. You could even use something like Workfabric in the middle to improve the training time, which is very effective. So this is how it is getting deployed as we speak.

So I want to talk quickly back on margin expansion, and we'll tie some of this back together. But -- on the 10 to 30 basis points of operating margin expansion, can you break down the key contributors across gross margin, SG&A leverage, utilization and even FX? Like how should we be thinking about the contribution of each of those elements?

Sure. So we have for '26 suggested 10 to 30 basis point margin expansion coming with a combination of both gross margin and SG&A leverage. Our message on gross margin is that of stability for the year, but it will have its own volatility through the quarters as we go. And the reason for that is some of the large deals when you win and when you're ramping them up, you tend to deploy cost ahead of the revenue generation. And therefore, you tend to manage that as a year, but quarterly, there could be variation around that. We definitely see contribution coming from an improved pyramid. We hired 20,000 recent college graduates in 2025. We're going to add to that number another 15%, 20% in '26. We will have a big thrust on the AI-led productivity, not just where we need to do as per the contract, but where we could improve the gross margin profile and retain with ourselves. SG&A, the leverage of AI on SG&A, right shoring of SG&A are another opportunity. So these are the levers with which we'll work through this.

So a big question we tend to get, obviously, and I'm sure you get it as well, is kind of a blanket question is one of pricing. And what are you seeing from a pricing perspective? I think some of your recent commentary was that maybe that was improving slightly, but I'm trying to measure that against things like gross margin and how you may be changing utilization and pyramid use, et cetera, to get to that flat gross margin that you're talking about.

Yes. So pricing on the new work related with AI is definitely better and superior to the current book of business. On the current book of business, definitely, there is a productivity pressure. I would say not pricing pressure. It's a reduction of unit of consumption versus the unit price reduction that you are looking at. And I think -- so this is the 2 words that we need to navigate that on new business, we price what is right, and we defend well on the existing pool of business.

So last few minutes here, I want to talk about the potential for India listing. I know that, that has been something that investors here in the U.S. have looked at with a great deal of interest, feeling like maybe that was a potential path to realize some value that maybe the stock and should have, et cetera. Any progress on that evaluation? And what are the main considerations and time line that investors should be prepared for as far as next updates and progress?

Absolutely. So I think the biggest biggest driver for it is that the -- there is a set of investors that would love to invest with us. And that pool is today not fully accessible by us. So listing that could make Cognizant as an investable stock by those -- that large pool of captive capital. Second is, of course, we are a large brand in India, and it could further enhance that, and our employees could own more local shares. The time line continues to be -- is something which is medium term. It is not immediate because there is a regulatory process. We continue to make good progress. But we are still not at a point where we have enough clarity regarding the decision of whether we go forward or we don't go forward. But we are definitely at a much -- we have far more awareness of the regulation and what it could entail compared to where we began the journey in end of October. So we continue to make progress, and we'll continue to update as we move forward.

So do you think -- obviously, there can be lots of considerations why you would move forward or not. But sitting here today, do you feel like it is feasible that if it is something that makes sense to you, et cetera, that can be done?

I think I would say we are progressing on that journey, and we don't have the view of the final regulation. So difficult to say yes or no to that question. But I would say economic rationale remains, and we'll work through the regulation.

So last 45 seconds here or so, Chad, what would be the primary thing that you would get -- have investors focus on and pay attention to for '26?

I would say it's just the role that companies like Cognizant Pen play and Cognizant has demonstrated it is how we capture the value in the new world of AI. I think that's our story, and we would request attention there.

Got it. Well, thank you very much. Thanks, everybody, for joining us, and have a good rest of the day.

Thank you very much, Jim. Thank you.

Ladies and gentlemen, welcome to the Cognizant Technology Solutions Year-end Fourth Quarter 2025 Earnings Conference Call. [Operator Instructions]

I would now like to turn the conference over to Mr. Tyler Scott, Senior Vice President, Investor Relations. Thank you. Please go ahead, sir.

Thank you, operator, and good morning, everyone. Welcome to Cognizant's Fourth Quarter and Full Year 2025 Earnings Call. I am joined today by Ravi Kumar, our CEO; and Jatin Dalal, our CFO. By now, you should have received a copy of the earnings release and the investor supplement. If you have not, copies are available on our website, cognizant.com.

Before we begin, I would like to remind you that some of the comments made on today's call and some of the responses to your questions may contain forward-looking statements. These statements are subject to the risks and uncertainties as described in the company's earnings release and other filings with the SEC. Additionally, during our call today, we will reference certain non-GAAP financial measures that we believe provide useful information for our investors. Reconciliations of non-GAAP financial measures where appropriate to the corresponding GAAP measures can be found in the company's earnings release and other filings with the SEC.

With that, over to you, Ravi.

Thank you, Tyler. Good morning, everyone. Thank you for joining us today. I'm pleased to report our momentum continued in the fourth quarter as revenue growth and adjusted operating margin again outpaced our expectations. Looking at the quarter's highlights. Revenue grew 3.8% year-over-year in constant currency, all organic, driven by North America. By segment, Financial Services led growth with constant currency revenue increasing 9% year-over-year during the quarter and 7% for the year, the highest annual level since 2016.

Q4 bookings grew 9% year-on-year, driving a record quarterly total contract value. We signed 12 large deals with a TCV of $100 million or greater, including deal valued at more than $1 billion. The value of these large deal wins is 60% greater than a year ago. Adjusted operating margin of 16% improved by 30 basis points year-over-year. We now have over 4,000 AI engagements across all three vectors and over 30% of our developer effort in software development cycles is AI-assisted and agentic. And our productivity improved as fixed bid and transaction-based work now represent more than 50% of our revenue.

We also saw a 5% and an 8% increase in trailing 12-month revenues and adjusted operating income per employee, respectively. These results drove 2025 revenues up 6.4% in constant currency, surpassing the $20 billion mark and the high end of our guidance range. Importantly, we delivered profitable growth. Our 15.8% adjusted operating margin exceeded guidance, rising 50 basis points over last year. We achieved this result while investing in our people, including through a merit cycle for most associates and our highest discretionary annual bonus funding level since 2018.

January marked my third anniversary as Cognizant's CEO. When we began this journey in early 2023, we set out to reclaim our winning heritage. In 2024, we successfully pivoted from stabilization to growth, industrialized our large deal engine and expanded our platform strategy with AI-led investments to broaden our capabilities. In early 2025, we laid out our strategic objectives to amplify talent, scale innovation and accelerate growth. We also set a goal to reach out industry's winners circle by 2027, and I'm extremely proud that we arrived two years early with top-tier revenue growth. Throughout 2025, we executed with speed and discipline, consistently meeting or beating the high end of our expectations each quarter as our investments began shaping Cognizant into an AI builder capable of scaling agentic AI across our clients' landscapes.

Looking at additional milestones that demonstrate our successful execution on our three strategic priorities. In 2025, we promoted more than 35,000 associates. We signed 28 deals, each with TCV above $100 million with a combined TCV up nearly 50% versus last year. This includes 5 mega deals with TCV of $500 million or greater. Our Net Promoter Scores reached a record high in 2025 from when I started three years ago. We expanded the breadth and depth of our partnerships across the hyperscaler and AI native landscapes. We signed and have since closed our acquisition of 3Cloud, adding more than 1,200 Azure specialists and engineers to industrialize our deep expertise in Azure, data and AI and application innovation.

We returned $2 billion to shareholders through dividends and share repurchases. Our progress is reflected in our total shareholder return, which was top two within our peer group in both 2025 and the three-year period beginning 2023 through 2025. Finally, with Belcan, we completed key integration milestones and continue to build a healthy synergy pipeline in the aerospace and defense industries. Last week, we announced Belcan secured a position on the Missile Defense Agency's Shield program. The indefinite -- delivery indefinite quantity contract with a ceiling value of $150 billion positions us to compete for a broad range of task orders supporting innovative defense capabilities.

As we enter 2026, our strategy is focused on solving the AI velocity gap, the gap between massive AI infrastructure spending in the past few years and business value realization for our clients. While AI technology is now mature enough to offer transformative value, the methodologies and tools to harness it are only just emerging and the value to enterprises hasn't drifted yet. In fact, our latest New Work, New World research released last month reveals that AI today is capable of unlocking $4.5 trillion in U.S. labor value in the future.

Cognizant's mission is to be the AI builder bridging this gap to enterprise value by converting the technology to measurable returns on investments for our clients. We are approaching this opportunity through our three-vector strategy. To capture Vector one demand, as we call it, we are applying AI-led productivity to augment and accelerate traditional software cycles. As we shared at our Investor Day, we see a massive multitrillion dollar opportunity to help clients accelerate the elimination of technology debt, build classical software in newer ways with AI platforms and repurpose savings towards innovation.

And to capture what we call vector 2 and 3, we are building entirely new cycles of agentic capital and digital labor that goes beyond the reach of legacy software, creating a much larger total addressable spend. Closing this velocity gap, the AI velocity gap requires new methodologies and evolving beyond the traditional IT services role of the last two decades. In the '90s, we were bespoke systems builders. We wrote custom software code, and we owned the outcomes. In the two decades that followed our role evolved into a system integrator.

We orchestrated classical software owned by various software providers. But classical software, which was written around the microprocessor was deterministic and built on rigid logic and fixed rules. Today's AI-led software, which is written around the frontier models is probabilistic and contextual. This shift allows us to own the stack again and deliver to outcomes. We believe reinvention and reimagination of businesses will be driven by value at the intersection of AI-led agentic capital and classical software.

To capture this demand, our AI builder stack acts as the connective tissue that addresses four layers of the ecosystem: AI compute, cloud, model access and human capital services. Let me share some key elements. First is our trademark basis framework, a proprietary blueprint that guides clients in architecting new business processes, specifically for deploying and orchestrating autonomous agents. This is a fundamental shift from writing rigid logic to designing behavior, persona, intent and outcomes. Second is our pioneering science of context engineering, a methodology for mapping a client's unique work graph, giving AI the situational awareness it needs to produce reliable business outcomes.

Context engineering bundles an organization's operating principles, tribal knowledge, work patterns, friction sources and historical and cultural imperatives so that AI intelligently binds to the enterprise's heterogeneous context, creating highly productive agentic capital. Third is our AI partnership ecosystem, which we continue to strengthen. On NVIDIA stack, we are offering solutions across the full life cycle from building and fine-tuning models to standing up agentic applications and deploying them as micro services. With Anthropic, Google Cloud, Microsoft Azure and OpenAI, we are using their frontier models and agentic tooling to build layers of application value to accelerate AI adoption for our clients.

With Adobe and Typeface, we are modernizing the enterprise marketing function and enabling cutting-edge customer experiences and content by moving manual workflows to Agentic orchestration. With Cloud Code Cognition, GitHub and Windsurf, we are industrializing software creation through advanced code generation. With Workfabric, we are scaling the emerging discipline of context engineering. With Writer and Uniphore, we are partnering to deploy specialized domain-specific AI platforms. With Palantir, we will integrate its foundry and artificial intelligence platform to support the integration of AI with our TriZetto business. And finally, with Salesforce and ServiceNow, we are embedding our Agentic networks directly into our clients' primary enterprise workflows.

The fourth layer of our AI builder stack is our own proprietary IP across platform services and research. For example, Flowsource elevates our engineering velocity, while Neuro ITOPS harnesses AI to proactively manage and self-heal hybrid environments. Our AI training data services have helped curate billions of high-precision data points for global clients. With TriZetto, we are accelerating and improving health care management. Our recently launched CareAdvance AI offerings help streamline clinical workflows, reduce administrative burden and empower care teams with faster and more accurate insights. And our award-winning AI Labs, which was awarded its 61st patent, continues to feed our continued investments in AI platforms and products.

To industrialize our AI builder stack, we have formed three units to sharpen our go-to-market muscle. First, our market-facing AI units are the hunters or value seekers working to capture the $4.5 trillion in labor value, our research identified. Second, our integrated AI solution unit acts as an architectural core, bringing various components of the AI stack together with strategic partnerships, cognizant methodologies and AI platforms to address specific reinvention needs of businesses. And finally, our centralized AI platforms and products unit is a factory packaging custom IP into repeatable solutions. Underpinning our AI builder stack is our talent strategy. Over the last 2.5 years, over 340,000 of our associates have completed AI skilling.

We are shifting from traditional linear staffing model to an asynchronous autonomous software engineering model. In this framework, our associates are trained to delegate complex high-value macro tasks to agentic networks while they micro steer to outcomes using platforms like Cognition, Gemini, Cloud, GitHub and others, orchestrating through Cognizant Flowsource. We are in the process of developing a hyper-productive, high-velocity delivery model for agents to asynchronously assist human software developers and agent managers. In addition, we are broadening our talent base with non-STEM talent and early career programs. This includes aggressively recruiting interdisciplinary skills at the intersection of industry domain and technology.

We added over 16,000 associates in India in 2025. In 2026, we are targeting 2,000 campus hires in the U.S. and approximately 20,000 in India. We are seeing this AI builder strategy translate into demand across our core practices. For example, our proprietary platforms like Flowsource and Neuro Engineering are helping clients unlock technology debt, helping to fuel 8% year-over-year in both the fourth quarter and year in our digital engineering practices. Similarly, our clients rethink their operations through an agentic lens, demand for our BPO business powered by deep immersion of digital labor grew 9% year-over-year in the quarter and the year.

Our AI data trading services launched early last year is gaining traction with our clients to build fine-tune AI models at speed and scale. And demand for data and cloud modernization remains healthy with revenue across both practices areas growing mid-single digits organically, outpacing total company growth. Now let me share a few client examples of our strategy in action. First, with a financial services client, we signed an incremental $1 billion partnership where we are leveraging our AI platforms, including our Neuro suite and Flowsource to help accelerate speed to market, drive product innovation and deliver enhanced productivity.

Next, with Cisco, the global leader in food distribution, we are transforming their complex customer interaction ecosystem into Agentic Capital. Previously, customer requests from product credits to order substitutions could have prolonged resolution window. Now by deploying orchestrated agents, we have collapsed that cycle to 90 seconds. Cisco is harvesting the AI-generated savings to fund its next phase of agentification. In the health care sector, we have moved from pilots to production-grade automation. For a major U.S. regional player, our AI intake platform reduced enrollment cycle times from as many as seven days to minutes. On the claims side, our clinical engine now adjudicates 96% of nurse note reviews autonomously, cutting human review times from 8 hours to 20 minutes.

We are scaling this expertise globally through a new strategic collaboration with Bupa Hong Kong, where our GenAI-led business process as a service solution modernizes claim and fraud, waste and abuse detection, marking our largest BPO win in the region. And we announced a multiyear expansion with Kohler, a leader in kitchen and bath products. Building on our successful 5-year partnership, we are bringing our cloud management capabilities and AI solutions like Neuro ITOPS to advance Kohler's digital ecosystem and drive AI-driven innovation. As we look towards 2026, we are well positioned to continue our momentum. Our ambition is to lead as an AI builder and maintain our position in our industry's winners circle.

In closing, I'm proud of all that we have accomplished over the last three years, which helped us reach our industry's winners circle two years ahead of plan. As the next decade of contextual computing unlocks new waves of nonlinear enterprise productivity and agentic software cycles, I believe there is a significant opportunity to create shared value for our clients, our associates and our shareholders. The foundation is set. I believe the boldest chapters of our story are still ahead. Thank you again for joining us. I'll now turn the call over to Jatin.

Thank you, Ravi, and thank you all for joining us. Our fourth quarter and full year results were a significant milestone in a multiyear journey marked by disciplined execution, strategic clarity and operational excellence. We delivered fourth quarter revenue growth above the high end of our guidance range and exceeded the initial full year guidance we provided in February last year across all metrics: revenue, adjusted operating income, EPS and free cash flow. We expect that our calendar year 2025 constant currency revenue growth will be in the top tier among the 10 peers against which we benchmark performance, placing us definitively in the winners' circle.

Beyond revenue growth, we achieved each of the broader objectives we provided at our Investor Day. This includes sustained large deal momentum, skilling for the future through AI training for approximately 260,000 employees, adjusted operating margin expansion of 50 basis points, and 2025 adjusted EPS growth of 11%, well above revenue growth. We delivered these results in a period of significant macroeconomic complexity and technological change, further bolstering our conviction in the strategic actions we have taken to become an AI builder company. We are well positioned to sustain this growth in 2026 and confident that we can build on this momentum in the years ahead.

Now moving to the details. In Q4, revenue of $5.3 billion grew 3.8% year-over-year in constant currency and was all organic. For the full year, the revenue of $21.1 billion grew 6.4% in constant currency, including 260 basis points of growth from Belcan. With respect to demand, the environment remains complex. Traditional discretionary spending cycles continue to evolve as clients rebaseline expectations for productivity gains; however, we view this as an opportunity to capture wallet share in large deals and help clients reinvest savings into innovation. Moreover, it opens new pools of addressable spend for us to advance our AI builder strategy.

Now turning to segment results. Financial Services once again led with full year constant currency revenue growth of approximately 7%. This was driven by strong performance in North America across banking, financial services and insurance clients. We have seen a steady improvement in discretionary spending in the last several quarters and consistent large deal signings, including a new mega deal in the fourth quarter. Our pipeline is strong, and we feel well positioned to carry this momentum in 2026. Health Sciences performance was resilient despite ongoing industry cost pressures and policy changes.

In this period of heightened uncertainty, we are helping customers reduce costs while improving patient experiences and accelerating productivity. These cost savings are funding clients' future-focused investment across core platforms, cloud modernization and regulatory readiness. We are seeing GenAI projects grow in areas like claims efficiency, clinical documentation and customer experience. And TriZetto remains a core differentiator, driving growth in implementation and managed services as clients modernize their administrative cores. Products and Resources performance has been stable. While tariff uncertainty continues to suppress discretionary spending, we expect large deal traction during the second half of 2025 to drive better performance in 2026.

AI adoption is growing across consumer and retail sectors, leading to demand for data services and Agentic-led experience transformation. In Communications, Media and Technology, fourth quarter year-over-year growth among our technology customers was more than offset by weakness in comms and media. Within Comms and Media, we have seen some impact from broader end market softness, particularly in North America. On the technology side, clients continue to rapidly innovate and adopt GenAI, which is driving demand for our services. Geographically, North America was again our standout region in the fourth quarter with growth of more than 4% year-over-year in constant currency, driven by financial services and health care. Europe grew 2% in constant currency with healthy growth in financial services and among life sciences customers. Rest of World grew in line with the total company, driven by the Middle East.

Turning to bookings. Bookings growth in fourth quarter was driven by robust large deal performance. We signed 12 deals each with TCV of more than $100 million. This includes two mega deals in the quarter, one in financial services and one in health care. On a trailing 12-month basis, bookings grew 5% and represented a book-to-bill of 1.3. Annual contract value declined modestly year-over-year due to the mix of longer duration deals and softness in small deal bands. That said, our backlog visibility at year-end is similar to where it stood this time last year and underpins our confidence in our full year guidance.

Now moving on to margins. Fourth quarter adjusted operating margin of 16% increased by 30 basis points year-over-year, benefiting from NextGen program savings, increased utilization and the Indian rupee depreciation. We delivered this result despite increased compensation costs, including our merit cycle and variable compensation, which drove a significant portion of our gross margin change year-over-year. Variable compensation for majority of our associates is expected to be the highest since 2018, and we remain committed to investing in talent to fuel our growth.

In November, the government of India implemented certain provisions of the code on Social Security or labor code as part of a broader labor law consolidation initiative. These rules did not have a material impact on our P&L in the quarter, but did result in a one-time increase to our defined benefit liability on our balance sheet with a corresponding increase to accumulated other comprehensive income. We anticipate a modest increase in our defined benefit costs prospectively.

Now to additional details on EPS, cash flow and capital allocation. Fourth quarter adjusted diluted EPS was $1.35, up 12% year-over-year. This drove full year EPS of $5.28, up 11% from the prior year. DSO of 81 days declined one day sequentially and increased three days year-over-year. Fourth quarter free cash flow was approximately $800 million and brought the full year amount to $2.7 billion, representing more than 100% of net income. During the fourth quarter, we returned nearly $500 million of capital to shareholders through share repurchases and dividends, bringing the full year total to approximately $2 billion.

We ended the quarter with cash and short-term investments of $1.9 billion or net cash of $1.3 billion. These amounts exclude about $730 million, which was deemed restricted cash and held in escrow ahead of the closing of the 3Cloud acquisition on January 1. Our M&A pipeline is healthy, and we intend to maintain an active acquisition strategy to strengthen our capabilities aligned with our AI builder strategy. We believe our robust free cash flow and strong balance sheet provide us with flexibility to invest strategically in the quarters ahead while continuing to return significant capital to shareholders.

Now turning to 2026 guidance. For the first quarter, we expect revenue to grow 2.7% to 4.2% year-over-year in constant currency. This includes approximately 100 basis points from our recently completed acquisition of 3Cloud. The midpoint of this range implies a modest sequential decline on an organic basis due to in part to lower bill days in Q1. For the full year, we expect revenue to grow 4% to 6.5% in constant currency. This includes inorganic contribution of approximately 150 basis points, of which approximately 1/3 is expected to come from future M&A. The midpoint of the range implies organic revenue growth of approximately 3.8%, which is consistent with our 2025 performance. This is also approximately 150 basis points above the midpoint of our initial 2025 organic growth guidance provided last year.

At the midpoint, our full year guidance implies stronger sequential growth in second and third quarter compared to 2025. And similar to our guidance philosophy last February, the midpoint is based on our current visibility and the discretionary demand environment as we see it today. Our adjusted operating margin guidance is 15.9% to 16.1%, which represents 10 to 30 basis points of expansion and is in line with the outlook we provided at our Investor Day last year. Similar to 2025, we expect expansion will be driven by the cost discipline and SG&A leverage. We expect free cash flow conversion of 90% to 100% of the net income. Adjusted effective tax rate is expected to be in the 25% to 26% range.

The midpoint implies a modest increase year-over-year driven in part by discrete beneficial items in 2025 that we do not expect to repeat in 2026. And our expected weighted average diluted share count is approximately 475 million. This leads to adjusted diluted EPS guidance of $5.56 to $5.70, representing 5% to 8% year-over-year growth. Expected EPS growth is being driven by anticipated revenue growth, margin expansion and lower share count. This is being partially offset by a higher tax rate, lower interest income as a result of lower assumed interest rates and an increase in nonoperating expenses related to the India Labor Code changes. For 2026, we expect to return approximately $1.6 billion of capital to shareholders, including approximately $1 billion towards share repurchases and the remainder towards our regular dividend. This leaves ample expected free cash flow available for future M&A.

As always, we will evaluate these plans regularly. In the absence of strategic and accretive acquisition targets, we expect to return capital to shareholders and not build cash on the balance sheet. Finally, as we mentioned last quarter, we continue to evaluate a potential primary offering and secondary listing in India. We have engaged various financial and legal advisers as well as the regulators in India to assess the idea. As always, we remain committed to acting in the best interest of our shareholders, and this process aligns with this commitment. As of today, the Board and management team continue to evaluate the proposal and have not yet made a decision. In summary, 2025 was a successful year. As we look towards 2026, we are well positioned to continue our momentum. As Ravi mentioned earlier, our ambition is to lead as an AI builder and maintain our position in our industry's winners circle.

With that, we will open up the call for your questions.

[Operator Instructions] Our first question comes from the line of Jason Kupferberg with Wells Fargo.

Nice to see these numbers. I just wanted to start on the AI topic and obviously, some new data points coming out from certain industry participants just over the last couple of days, for example, talking about expediting ERP implementations pretty significantly. It certainly seems to us like Cognizant to date has been a net winner from AI. I wanted to get your perspective on how that plays out in '26 and maybe just in light of some of these recent headlines, what percent of your total revenue currently comes from package implementation?

Thank you for that question. This has happened over tech revolutions before. When a new technology comes, we kind of think the old technology will go away, but the new technology will actually provide more opportunities. I see this as an increase in our total addressable spend, I mean if you're referring to what's happening in the last few days, I can tell you any tool, any technology will magically not generate value on the other side. You need a bridge, and that bridge is what companies like Cognizant do. And I'm going to be precise on what I mean. You can't apply this technology on existing old processes. So you have to reinvent and reimagine a process.

This is a technology which is very contextual in nature. It's written over -- it's not written on the microprocessor, which is deterministic. It's very probabilistic, which means we have pioneered something called context engineering, which is grounding this technology in the reality of an enterprise, understanding the heterogeneity of an enterprise. I mean, the two SAP implementations are not the same, just to go back to package work you spoke about. And it is about understanding the hassle, the flows and everything else, integrating deterministic software, which was written for the last 25 years with probabilistic software, which will be written for the next 25 years and building flows where digital and human labor can work together, integrating it into the operating and the physical layers of an enterprise.

I think all of this is a lot of heavy lift. I mean if this was all real and if this was -- it would have switched on magically without anybody doing anything, we would have seen the drift of value already, and that's not happened yet. There is -- our study said is $4.5 trillion of labor, which can actually be amplified with higher productivity out of the $15 trillion in the United States in the last few years. It's not drifted yet because all of this has to be done. Equally, going back to what you just asked, there is technical debt. There is a lot of backlog. There is the elasticity of software, the traditional software, leaving all the new software we're going to write, which can actually expand.

So we see this as a net new tailwind for us on two swim lanes on the traditional software, apply it and do more for less and get more consumption because of elasticity, take out technical debt, take out the backlog. On the other end, apply this on a much new addressable spend, which classical software didn't penetrate. So I see this as more of a bigger opportunity for us and a higher -- with a higher surface area for us to actually operate. So this is a tailwind. We are turning out to be winners. Our builder strategy is working.

And our three vector strategy we spoke about, both on applying this to traditional software and writing new agentic software, which can actually capture significantly more surface area and enterprises. We think it's a phenomenal opportunity. Now enterprise software package, which you spoke about, package software has been there for the last 20 years. There has been deterministic code. There's been systems of record in it. We're going to apply layers of AI value on top of it, actually generate more value than before. So there is going to be a coexistence of deterministic and probabilistic software, and there's going to be interplay between the two.

Okay. Understood. And just a numbers one for Jatin. I wanted to ask about gross margins. It sounded like the year-over-year decline in Q4 was primarily due to higher variable comp, which is arguably a good problem to have, just given the overall financial performance of the company this year. But any other gross margin dynamics we should be thinking about in terms of 2026? Do you expect gross margins to be up year-over-year? And just to clarify, are you seeing any like-for-like pricing pressure as part of the year-over-year declines in gross margin currently?

Sure. So yes, the Q4 impact on gross margin was predominantly on account of higher bonus funding that we did for the full year, in quarter 4, led by a strong operating margin performance for the full year that we were able to deliver. Apart from that, there was also a salary increase, as you are aware, which came through effective 1st November into the gross margin. So I would say those are the two factors that played out a bit in quarter 4. I think the right way to see our gross margin is for the full year. And the full year impacts are predominantly, one, the Belcan impact for the full year in 2025 gross margins. And the second is essentially slightly -- I mean, essentially the higher bonus for 2025. And as you mentioned, that's a good thing to have.

Looking ahead for 2026, we -- I mean, there is a productivity-led pressure in the industry. And therefore, the expectation that for the -- for a dollar value, you get a superior throughput than what you have traditionally enjoyed through traditional productivity levers in past. And that does impact the revenue, but I wouldn't call it sort of a -- I wouldn't call it a drag on margins yet so long as you are able to execute on your internal productivity measures and keep the cost curve below the price curve, I mean, continuously. And therefore, you have seen we have been able to deliver revenue per employee productivity and profit per employee productivity in previous 12 months. So, so far, we have been able to execute well against that market momentum for productivity. And therefore, I would say we are entering 2026 with that confidence.

Going forward, we'll have to continue to watch out for the movement in the market. We do think that we have a few levers apart from AI productivity and a couple of them are really the continued improvement in pyramid. We hired 20,000 fresh college graduates in 2025, and we'll continue to look at that. And that does impact a long-term cost structure of the organization. And we'll continue to look at other traditional measures like offshoring and utilization beyond AI productivity that I spoke about. So overall, we have things we can work through for 2026.

Our next question comes from the line of Tien-Tsin Huang with JPMorgan.

Really strong large deal activity again here in the fourth quarter. So I want to ask your confidence in your ability to grow off of that larger base in '26 over '25. How does the pipeline look for larger deals in '26? And any good line of sight into deal ramps being timely?

Thank you, Tien-Tsin, for that question. Yes, I think we've had a great bookings quarter, 9% Y-o-Y, TTM 5%, 50% increase in TCV on large deals for the full year and 60% increase in TCV of large deals in quarter 4, and we are very excited about the fact that our fixed price business now is almost 50%. I mean three years ago, that used to be 41% to 42%. So we can, in some ways, fixed price it, share the productivity with our clients and actually pass on some of it to ourselves. We are keeping that. We are one of the -- we are probably the only player in our peer group, which talks about code- assisted and autonomous software engineering. 32% of our code is AI-assisted.

So we have now activated two swim lanes. In 2024, a lot of the large deals were productivity led. Now we are seeing innovation-led, vector 2, vector 3, as I call it. We did twelve $100 million deals in Q4. So it was a record of starts. We crossed $10 billion. That's a record of starts. We have one $1 billion deal in quarter 4. We have five mega deals in the full year, and we have two mega deals in quarter 4. So we have a strong pipeline, and we have activated both the swim lanes, and we are starting to do transition of that work. And therefore, we see a solid quarter 2 and quarter 3. In fact, we see more acceleration during the year of ramp-up as well as more deals on the way. So I'm very excited about the fact that this has become a tailwind for us. AI is a tailwind for us.

No, terrific. It's impressive. And so just to clarify, you mentioned it there, Ravi, or Jatin, if you want to chime in. Just the confidence in the faster sequential growth beyond the first quarter being higher than the pattern in the last couple of years. So it sounds like that's really just what you see in terms of the large deals ramping and the timeliness of that.

Sure. So, there are two factors at play there. One, of course, is a strong bookings that we are walking in 2026 with. And the second is there is some amount of seasonality between quarters also in 2026 compared to '25. And for example, in Q1, there are lower bill days in '26 compared to the number of bill days that we had in Q1 in 2025, which automatically means that the sequential number improves in quarter 2 compared to quarter 1 in 2026. So these are the two factors that give us confidence that we can execute better sequential growth in the middle of the year, and that's what we have assumed in our guidance range, including the ramp-up of deals which we have closed in quarter 4.

Our next question comes from the line of Keith Bachman with BMO Capital Markets.

I wanted to ask about the risk and opportunities of the fixed price or success-based contracts that are now about 50% of total. And what I'm trying to understand is you're pricing -- I think you're pricing these contracts on assumed cost curves that leverage new innovations, including AI. And I just want to understand how -- A, is there more -- are these -- are the prices more aggressive today than they have been? B, how should investors think about the risk and the opportunities of overs and unders in terms of achieving those cost curves? In other words, are you sharing those risks with the customers? But if you could just speak to the changing nature of the economic risks associated with these success-based contracts?

Sure. So by -- there are various types of fixed price engagements. But essentially, I mean, they have one thing in common is that the larger component of delivery risk resides with the service providers like us. And essentially, we underwrite a productivity in the beginning of the contract and we deliver to that productivity to the customer irrespective of whether we are able to achieve that outcome from a cost standpoint or not. The history of industry is that we always have found ways through new technological progress to be able to deliver it. Specifically in light of the whole large deal momentum that Cognizant has been able to achieve.

We have a very, I would say, very robust process of bid versus bid that we monitor every month, the performance of the deals that we have won and how we are delivering our operating margin and revenue performance against those promises. And I'm happy to share that we deliver on aggregate of the portfolio very close to the expected margins that we had planned, which means we are in aggregate, not having any overrun or also not significant underrun. So overall, we are tracking to the budgeted goal for our customers as we go. And that we will continue to do. There is something that is crucial in times like this when technology is shifting. And overall, we feel we are performing well.

I just want to add two quick points to Jatin. If you look at it, in some ways, we are sharing the productivity, sharing the risk with our clients, but we are actually doubling down on execution. Look at our revenue per person and margin per person. It's gone up by 5% trailing 12 months, 8% margin and 5% revenue trailing 12 months, which essentially means we are able to share with our clients the productivity, win, actually price to win and deliver to margins. That's our motto.

And I think with this nonlinear opportunity with the technology, you can kind of be ahead of the curve and do that. The second, I would believe, which is a very important shift. Historically, if you look at it, go back to the '90s, companies like ours used to own the outcomes, and we used to price on outcomes. And then the enterprise software either both on-prem and SaaS and then the plumbing on the cloud kind of abstracted layers of that value. And outcome-based was hard because there are so many people in the mix.

Here, we are fast forward. We can own the outcomes. We can own the outcomes. We can make this a platform play. We can make it nonlinear cost and nonlinear revenue. And we can take over operations of companies and give them a service, which -- that is the reason why our BPO business is actually growing at 9% Y-o-Y. 9% of the BPO business growing because we are able to do that very well. We are able to deliver to outcomes, own the value chain and share the benefits.

Ravi. This sort of led into my next question was durability of BPO. I think two years ago, many, including ourselves, had some concerns about what AI would do to BPO. It's been, I think, one of the more robust parts of the market. And it seems clients need help in setting AI into BPO. And my question is, how durable is this? In other words, once you get those processes established in BPO enabled by AI, does that create longer-term headwinds? Or is there enough momentum here that this is a multiyear tailwind or good growth within the context of BPO?

That's a great question. I mean, look, this is a total addressable spend, which is 10x or maybe 20x more than tech spend because you're embedding technology data into process and in recent times to machine learning and AI. Cognizant has had 9% to 10% growth in BPO for three years in a row. And the reason why we have done so is because we have always been on the cutting edge. We think this is a long-ish tailwind because operations of companies is a much bigger addressable spend. And we think we have an opportunity not just to transform, reinvent, reimagine flows in a company, we also have the ability to maintain them. I think we are underestimating how much that reinvention will need. It's decades of work.

We are underestimating how much it needs to maintain. I mean this is a contextual technology. It has to be grounded. It has to be situational. And the effort needed to maintain and manage deterministic technology is less than the effort needed to maintain a probabilistic technology. So we have actually more work to do in maintaining than before. So I see this as a significant tailwind to our BPO business. We call it intuitive operations even before AI came into picture. So that's how we see this.

Our next question comes from the line of Jim Schneider with Goldman Sachs.

You talked on the Health Sciences script about the cost benefits to those companies sort of outweighing any kind of regulatory pressures you're seeing. Clearly, in the payer space, there's been a lot of debate about additional regulatory burdens and cost pressures. Just would love to understand your level of confidence in the relative growth for your Health Sciences segment this year relative to your full year guidance overall.

Thank you. Our Health Sciences business grew at 6-plus percent, way higher than our company average. It's a business where we probably are the #1 player in the market. We have a platform with $0.5 trillion of transactions flowing through it. We have 200 million members. We have a moat which is super differentiated. There is a lot of labor sitting around it. And I think with the uncertainty of regulation in the payer side, you will want to transform those layers of value around the TriZetto business and shift that money to care because there is uncertainty around spend cycles.

So we are seeing more traction with companies which are willing to apply agentic in and around the TriZetto platform. And we see this capture of these value pools a new spend area for Cognizant. We have started to partner with Palantir, which I spoke about. We have a partnership with Microsoft. We have a partnership with AWS. We are doing work with Google Cloud. We're putting all these layers, and we are agentifying the labor attached to it so that the administrative costs are going to go down, and that money is going to be underwritten for care. So there is more hustle and more work because of the uncertainty and the need and the paranoia about transformation so that this money is going to be moved to care.

Equally, there is a lot of work around applying agentic to, say, bedside care or applying agentic to the life cycle of patients all the way from before they start to get to a doctor to after they finish the visit to the doctor. In fact, some of the places I've mentioned one or two examples where we are able to take notes of doctors and nurses and agentify the whole thing and create productivity, high productivity for health care workers. So I see this as a tailwind because of the fact that there is uncertainty around regulation, we are actually going to see more transformation on the administrative layers, which will then transfer that value to care.

Of course, there's also a part of life sciences and providers ecosystems there. And remember, the regulatory pressure is only on Medicaid and Medicare. It's not on commercial health care. But having said that, I think that uncertainty provides an opportunity to constantly innovate and transform and also to adhere to new regulatory norms using technology to adhere to new regulatory norms.

And then maybe as a follow-up, you sort of discussed many things that are sort of impacting gross margins at this point, whether that be the kind of outcome-based pricing, the fixed pricing and also the pyramid. Can you maybe talk about when -- do you see line of sight to sort of gross margin inflecting on a year-over-year basis at some point during the course of this year?

Yes. So Jim, we -- I mean, we have guided for the overall operating margin line. I don't want to guide at both the lines, but our endeavor would be to strive to reach that improvement in gross margin line, too. As I shared before, 2025 doesn't worry me because I know this -- our core margins have remained protected. The dilution that you see in 2025 is coming predominantly because of Belcan, which is structurally more on-site-centric work. And therefore, it is not about lower profitability. But since you add a business which is more on-site centric, it is bound to have a lower gross margin, as you know.

So it is not -- it is just a portfolio which has a particular characteristics, which have got added to the larger portfolio. That's one reason. And the second reason is really the higher bonus payout, which I think is a good thing for our employees. So I know we have protected and sustained the margin in '25. We will work towards, of course, improving them in future...

I just want to add two quick things here. Look, we are broadening the pyramid. We have a thesis that the value is actually going to be more at the bottom with higher productivity. Last year, we added more school graduates than the previous year. This year, we're going to add more school graduates than the previous year. So that's going to give a tailwind to it. Our productivity sharing with our clients and how much we are going to keep back, which is the 5% revenue per person and 8% revenue margin per person, that's going to help. And good discipline, operating discipline has also helped. So there is tailwind on it, and we are not worried about keeping our expansive margins for 2026 and beyond.

Our next question comes from the line of Bryan Bergin with TD Cowen.

I wanted to start with just kind of a bookings to growth question. So can you help bridge the ACV growth performance in '25 to your '26 growth guide? Just curious how you're factoring things like pipeline conversion and really at the midpoint of the range as well as things like short-term work. Can you detail that first? And then I'll ask my follow-up here. On the margin front, just SG&A, you've driven meaningful savings here in '25. You've actually held the dollar level flat for like three years. Understanding it wasn't optimized before. I'm just thinking how much more meaningful room do you have in that SG&A line to continue to help you?

Yes. So on ACV, definitely, we saw some amount of softness in quarter 4, but I would also characterize that with the bundling of smaller deals being given out as consolidated contracts. And therefore, you see a significant increase in TCV of large deals which corresponds to that shrinkage for the smaller deals, and that's a sort of an industry dynamic that we see in times like this. So while that is definitely a data point, but it is not something that is a challenge from a growth standpoint. On your -- sorry, can you just repeat your second question?

Yes. Just on SG&A. So you've done a great job there, right, for a couple of years. I'm just curious how much more room you have to optimize that base to continue to help if gross margin isn't going to stabilize sooner?

SG&A continues to be an area of focus for us. So while we have done a good job in '25 and '24, so two years in a row, but that has now additional opportunity in form of the deployment of AI in the corporate work that we do. So certainly, we will continue to push that in 2026, too.

Our next question comes from the line of James Faucette with Morgan Stanley.

Just wanted to ask a couple of quick follow-up questions. On near-term activity and sales engagement, I think you've mentioned a little bit of softness there, at least in the fourth quarter. Can you just give a little more color there? Where were you seeing that? Is it just in near-term bookings? And how are you feeling about the potential for discretionary work to come back? I know that, that's been something that has been -- everybody has been looking forward to coming back a little bit more aggressively. And clearly, you're doing well in kind of the larger deals, but just wondering about kind of more of this faster turn business.

Yes. So look, we'll continue to see more large deal momentum. I mean, if you want to share productivity with clients and win and use the process of consolidation, wallet share swap, that's a phenomenal opportunity. Innovation levers are starting to kick in now. That's going to help us on smaller deals and discretionary. Look at financial services. We did 9-plus percent quarter 4, and we did 7-plus percent for the full year. Financial services is a lot of discretionary. So -- and it's our largest vertical. And financial services performance in 2025 is the best we have had since 2018. So it is actually a good tailwind.

I think there's going to be -- as the pivot for AI shifts significantly from productivity to innovation, we're going to see more discretionary flowing in. There is a talk of physical AI, which is now starting to hit manufacturing and automotive and aerospace and industries of that kind. That will also create opportunity. As the AI experiments will start to go into production, you're going to see discretionary new value pools opening up. So overall, I think financial services is a positive news, and it's one of our most cutting-edge industries and the highest exposure. So -- the others will follow. So I do see the unlock. Of course, the macro has to support for the acceleration, for discretionary macro has to support.

What I'm not worried about is actually, I would say, if the AI advances have to trickle drift to the businesses, I would actually believe that, that is actually going to support the discretionary to come back. It's going to be a catalyst. It's going to trigger a CapEx cycle on enterprises to drift that value, and it will flow through to us. So that's how I'm seeing it. So financial services is the starting point, which has already happened. The others will follow.

Yes. Thanks for pointing out financial services. That stood out to us as well. And then just quickly, I know you gave a quick summary of the work that you're doing exploring the India listing. Can you just give us a rough idea of what you're thinking about in terms of time frame or when at least we should be able to put together a calendar and time frame list? I know that's a key concern or at least thought for a lot of investors.

Yes. So as I mentioned in the opening remarks, we continue to make progress. We are engaged with our advisers. At this juncture, we are still thinking through the decision, the regulatory framework and therefore, the decision around the imminent secondary -- sorry, primary offering and secondary listing. And -- at some point, we should be able to come back and tell you more about this. But at this juncture, I think it's a continued progress would be what I would suggest as an update from previous quarter to this quarter.

And the -- and we have had constructive discussions with the regulators. So we're continuing to do what is right for our shareholders and continue to look for more investors to be a part of our growth story. So we'll keep you updated on that.

Maybe we'll take one more.

Our final question today comes from the line of Rod Bourgeois with Deep Dive Equity Research.

Okay. Great. And I'll just ask one given the time here. You already addressed the question about AI being applied to ERP implementation. There's also new cloud plug-ins geared towards workflow automation. So I wanted to ask to what extent you see such workflow automation abilities impacting your market opportunity? And to what extent you already have a partnership also in that area?

Thank you, Rod. Look, we announced a partnership with Anthropic last year -- late last year. The more AI can do, the more is the opportunity for us. It's very simple. I mean, let's take about the plug into legal. How much software has been implemented in legal. There is so much paralegal work happening. In fact, we work for a professional services and legal services company where we are agentifying all that paralegal work. That never existed before. Now there's a lot of labor around classical software. And all that labor needs more productivity. If you want to drift that value to higher productivity to the workers in every function of a company, AI can be the catalyst. And that is net new spend area for us. And that is what we are looking for. Those net new spend areas. This is a totally new addressable spend.

And if you're embedding technology, you are able to integrate this technology to the systems of record written and SaaS software and you're able to build those workflows, build those flows where humans and digital labor can work together and amplify the productivity. If you're able to reinvent those processes for higher throughput, higher velocity, using this tool as a catalyst, we're able to bump up the productivity of enterprises and bump up the productivity of workers, and we're going to be the bridge to do that.

So I see this as a unique new opportunity. The more comes in. I mean, remember, this technology is smart enough already. I mentioned in our remarks that $4.5 trillion of labor can already -- is already exposed to AI, and it can -- you can create higher productivity. The reality is none of that has drifted to enterprises. None of that has drifted to enterprises. You need that bridge. And the bridge is all about contextual engineering. It is about reinventing the process. It's about redesigning these flows in a company, integrating it into the SaaS layer so that there is interplay between deterministic and probabilistic layers. And it is also about integrating it into the physical and the operational layers of the company so that you can get that value. So that is the way forward.

And we don't -- I mean, at this point of time, it is not about getting the smarter technologies. We already have smarter technologies. At this point of time, how do you drift that value to businesses -- and there is an urgency because there's $400 billion to $500 billion, which has been spent on infrastructure in the last two years. And you have to get that value here and there's trillions of dollars of value and the shelf life of this technology is short. So I see this as a net positive for more work, more surface area, more addressable spend for companies like ours, and we call it the AI builder because we have this unique opportunity to be the bridge.

So thank you so much for joining the call. Thank you for your continued support. We've had an exciting 2025. We have outpaced our own expectations on revenue, margin, EPS.

EPS growth has been higher than revenue growth, expansive margins. This is what we said in the Investor Day and we are continuing to keep our trajectory -- accelerating our trajectory. We are on the top of our charts on our relative growth in comparison to our peers. And we hope to keep the winners circle performance in 2026, expansive margins and EPS higher than revenue growth and revenue growth at the middle of our range is actually higher than what we presented last year. And we are in a solid foundation. And I think the boldest chapters are going to be in '26, '27 as we go forward.

Thank you. This concludes today's Cognizant Technology Solutions Year-end Fourth Quarter 2025 Earnings Conference Call. You may now disconnect your lines. Thank you for your participation.

Ladies and gentlemen, welcome to the Cognizant Technology Solutions Third Quarter 2025 Earnings Conference Call. [Operator Instructions].

I would now like to turn the conference over to Mr. Tyler Scott, Vice President, Investor Relations. Please go ahead, sir.

Thank you, operator, and good morning, everyone. Welcome to Cognizant's Third Quarter 2025 Earnings Call. I'm joined today by Ravi Kumar, Chief Executive Officer; and Jatin Dalal, Chief Financial Officer.

By now, you should have received a copy of the earnings release and investor supplement for our third quarter results. If you have not, copies are available on our website, cognizant.com. Before we begin, I would like to remind you that some of the comments made on today's call and some of the responses to your questions may contain forward-looking statements. These statements are subject to the risks and uncertainties as described in the company's earnings release and other filings with the SEC. Additionally, during our call today, we will reference certain non-GAAP financial measures that we believe provide useful information to our investors. Reconciliations of non-GAAP financial measures where appropriate to the corresponding GAAP measures can be found in the company's earnings release and other filings with the SEC.

With that, over to you, Ravi.

Thank you, Tyler, and good morning, everyone. Thank you for joining us. We are pleased to report another industry-leading performance in the third quarter of 2025 as revenue growth and adjusted operating margin again outpaced our expectations. Our results reflect the momentum we have built over the past 2.5 years, helping clients embrace AI. Our investments in platforms, intellectual property, partnerships and in upskilling our people are evolving Cognizant into an AI builder capable of scaling agentic AI across the enterprise. As AI infrastructure expands, our clients increasingly need support from partners who can help them move from experimentation to enterprise-wide adoption with speed, precision and trust.

Turning to third quarter highlights. Revenue grew 6.5% year-over-year in constant currency to $5.4 billion. All 4 of our operating segments grew revenue organically year-over-year. This breadth of performance across industries and geographies reflects the strength and resilience of our portfolio capabilities and delivery model. This is the fifth consecutive quarter of year-over-year organic revenue growth. Our strongest sequential organic growth since 2022 and another podium finish to our peer group's Winner Circle.

We signed 6 large deals, each with TCV of $100 million or more, bringing our year-to-date total to 16. Trailing 12 months bookings is up 5% year-over-year and year-to-date the TCV of our large deals is up from the prior year period. We are focused on converting value from AI across our 3,500 early AI engagements and embracing AI in the delivery of our services and to drive internal transformation. As we do this, we are also increasing our fixed bid transaction and outcome-based services mix. And we are beginning to see trends of nonlinearity emerge.

For example, on a trailing 12-month basis, revenue per employee rose 8% year-over-year, while adjusting operating margin income per employee grew 10%. As we continue to scale our IP and platforms, we expect more examples of nonlinear AI-led growth to emerge.

Importantly, we are expanding margins while continuing to fund our organic inorganic growth initiatives and increasing returns to shareholders. Q3 adjusted operating margin improved 70 basis points year-over-year, driven by our disciplined expense management, along with our increasingly AI-enabled delivery model. Our year-to-date performance has put us on track to outperform the revenue guidance we established at the beginning of the year, and we expect to meet the high end of the adjusted operating margin range we set then.

For much of the last 30 years, IT services grew through a linear model. More people and more projects drove incremental growth. AI is reshaping that equation by compressing time, cost and complexity and redefining how value is created. The opportunity to partner with the clients and drive outcomes is now more expansive, immersive and elastic. The progress we are sharing today reflects 2.5 years of focused execution and amplifying talent, scaling innovation and accelerating growth to return Cognizant to a leadership position in the AI era.

Becoming an AI builder means building the platforms and engineering capabilities that enable agentic AI to scale across the enterprise. Our progress begins with our workforce as we enable AI fluency across the 350,000 associates. We continue to fuel strength and future readiness for our associates through a learning engine and access to AI tools, which is why we are hiring more new graduates across the world this year and investing in their AI upskilling. In July, our Vibe Coding initiative earned Cognizant the Guinness World Records title for the world's largest online Generative AI Hackathon. More than 53,000 associates across 40 countries built over 30,000 working prototypes, improving their AI code assist skills and productivity. And we are continuing to expand into the AI ecosystem.

Recently, we entered a new collaboration with Anthropic. Under our agreement, we plan to deploy Anthropic's cloud models and agentic tooling with our platforms to help clients scale AI, while also deploying them internally to advance our own operations. Our AI Builder strategy is anchored in 3 distinct vectors. AI-led productivity, industrializing AI, and identifying the enterprise. When each vector is advancing at a different velocity, together they're performing a flywheel of new value creation. Let me provide an update on vector 1. AI-led productivity is the funding engine for enterprise transformation as we help clients accelerate software development, lower deployment costs and reduce technical debt that we estimate is costing enterprises hundreds of billions of dollars in annual servicing.

In the third quarter, approximately 30% of our internal code was AI generated, significantly improving productivity of our developers. We believe it could reach 50% in the years ahead. A great example to illustrate our client impact with Code Assist platform partners is a recent award as the AI GitHub Services and Channel Partner of the Year in recognition of our achievements in helping clients with our AI transformation initiatives. Many clients have asked us for support in bringing wipe coding and code assist best practices to their organizations.

We recently launched a Cognizant enterprise wipe coding blueprint, bringing our playbooks and insights to clients seeking to build AI fluency across their own teams. This transformation extends beyond the developer community. Internally, we have embedded AI across more than 150 use cases from finance and operations to sales enablement and contract pricing. These applications are streamlining decision-making, improving accuracy and accelerating cycle times.

A primary tool for executing Vector 1 is our Flowsource platform, which integrates generative and agentic AI across the full software development life cycle. Flowsource is now being used at over 70 clients with an additional 120 in the pipeline. One of those clients is Pearson, where we are using AI and digital technologies to modernize their learning platforms, products and applications by leveraging Flowsource.

Our proactive shift to AI native and platform-driven engineering accelerates the software development cycle by enabling engineers to deliver enterprise-grade AI-infused digital applications with greater speed and scale. This is showing up in our results with our approximately $2 billion annual run rate digital engineering business growing about 8% organically year-to-date. Vector 1 is also fueling our large deal momentum as clients consolidate their software estates and shift to outcome-based models, they're capturing savings and unlocking higher value. Often reinvesting those gains into Vector 2 and Vector 3 initiatives. It is creating a self-reinforcing cycle of transformation.

A great example of this in action is our cloud and infrastructure modernization business, which grew 10% year-over-year in the quarter. Our AI tooling and services in the space has helped over 25 clients so far to build, respond and resolve to reliant and resilient IT infrastructure. Now more on vector 2 or industrializing AI as the scalability layer. It's about moving AI beyond experimentation into enterprise-grade systems, building AI-ready infrastructure, integrating contextual data and operationalizing AI responsibly.

It also involves developing new business operating models, leading to an interplay of software and agentic layers, human and agentic capital and structured and unstructured data to reimagine an enterprise. We are leading this effort with our consulting basis framework and methodology to help clients reimagine business processes as they develop and deploy agents. And we are deepening our expertise with the next level capability set, including Agent Foundry, a framework and library of the industry and workflow-specific agents, helping power agentic AI at scale. Together with our clients, we have developed more than 1,500 agents across the company.

Second, AI data training services, where we have over 10,000 specialists fine-tuning models with domain-specific context. We have supported leading tech companies with training their machine learning systems long before generative AI entered the mainstream, and we are now bringing the same expertise to Global 2000 clients. Third, small language models development. Fourth, context engineering, which we believe is one of the most critical emerging disciplines in enterprise AI to capture enterprise workflows, domain and tribal knowledge, personas, rules and execution patterns. It is the connectivity tissue between models and outcomes.

In partnership with Work Fabric AI, we are deploying context engineers who are helping clients build tailored foundations for AI adoption. And finally, IP on the edge, which I began describing last quarter is a horizontal foundation layer where we are bundling platforms like Neuro AI with services and IP to deliver outcomes. With 400 platform deployments already in motion, we are helping clients modernize core systems to reduce risk, accelerate time to market and improve experiences. As we build layers of contextual value on foundation models through a combination of context engineering, SLMs and multi-agent systems, we are delivering numerous production-grade AI use cases.

To bring this to life, we helped the national grocery chain optimize it in-store pickup process for online orders, reducing fulfillment time by 20% to 45% through smarter inventory selection, product substitutions and routing. This is driving a measurable increase in online orders. Lastly, vector 3 or identifying the enterprise is about unlocking exponential agentic capital. Historically, we built software for humans.

With Agentic AI, we now reimagine processes end-to-end by deploying agents with humans in the loop to deliver outcomes. This expands the enterprise's surface area, enabling a blended human plus agent workforce across new domains. The agentic development life cycle or ADLC differs fundamentally from the traditional software development cycle or SDLC. SDLC is structured and deterministic, input in, output out. ADLC is adaptive and outcome-driven, to design for behavior, supervise performance and evolves capabilities over time.

We believe ADLC significantly expands our addressable market, demanding deep ownership to manage human digital collaboration. As an AI builder, we are creating an agentic ecosystem where agents reason, adopt and collaborate, unlocking service capabilities that weren't possible before.

Cognizant is an early launch partner for Google Gemini Enterprise, an AI-powered platform designed for enterprises to drive unified secure AI capabilities. It seamlessly connects enterprise data, tools and workflows and leverages Gemini models to enable agentic journeys. And some client examples include reducing order response times from 5 days to 90 seconds with digital sales agents for a leading food distributor, helping a leading provider of cell-free DNA diagnostics reinvent patient education, access and onboarding processes, modernizing order management for a crop sciences company using Agent force, delivering intelligent lead generation for a top labeling and a packaging provider.

With TriZetto's core adjudication platform supporting health plans, we have deployed multi-agent workflows that connect TriZetto agents to front-end experience platforms such as Salesforce, Genesys and ServiceNow to address common interactions such as requesting ID cards from a member or giving provider the status of a claim. We believe much of the Vector 3 will flow into Intuitive Operations and automation practice, which is our BPO business, including BPaaS services.

Our BPO revenue grew 10% in the last 2 quarters and is on track to reach $3 billion in annualized revenue over the next several quarters. We believe identification will unlock new labor pools, including roles that don't yet exist. As digital labor diffuses into enterprise operations, the nature of human endeavor will evolve. Together, our work across vector 1, 2 and 3 reflects our evolution into an AI builder company, one that blends deep domain expertise with platform innovation and interdisciplinary talent.

30 years ago, IT services companies were builders, crafting the foundational systems that powered industries. Over time, the role shifted towards integration, development and maintenance and growth became more linear. Today, Cognizant has a unique opportunity to reclaim the builder mindset and capture a greater share of the fragmented AI market.

The scale of this opportunity is extraordinary. While global software market is in hundreds of billions of dollars, the surrounding labor spend represents many trillions more. Classical software has barely penetrated that space. We believe AI's winners will be those who diffuse into this labor spend, reshape how work gets done. Software and agent development cycles will coexist, and Cognizant is poised to generate layers of value in this expansive new role for enterprise reimagination.

In closing, we are proud of our Q3 results and the momentum we are building financially, commercially and strategically. We are evolving from software implementer to AI builder powered by an engineering heritage and AI-ready workforce and proprietary innovation. We know long-term success will be determined by the outcomes we deliver for our clients, our people and our shareholders.

Thank you again for joining us. I'll now turn the call over to Jatin.

Thank you, Ravi, and thank you all for joining us. We are pleased to report third quarter results that include revenue growth above the high end of our guidance range, strong margin expansion year-over-year and double-digit adjusted EPS growth. Our performance once again places us in the winner circle, and we are delivering these results despite a complex demand environment and geopolitical backdrop.

We continue to execute with discipline, driving improved revenue growth while investing in our people, technology and partnership to support our AI builder strategy and long-term growth. At the same time, we are delivering consistent margin expansion. These results are underpinned by balanced capital allocation framework, which we believe are key enablers to driving long-term and sustained shareholder value creation.

Now moving to the details of the quarter. In Q3, we delivered revenue of $5.4 billion, up 6.5% year-over-year in constant currency, again led by strong growth in North America. Belcan contributed slightly less than 250 basis points of inorganic growth. Year-to-date, our revenue grew 7.3% in constant currency, including 350 basis points of inorganic growth. Adjusted operating margin expanded 50 basis points and adjusted EPS grew approximately 11%. And we returned about $1.5 billion of capital to shareholders.

With respect to demand environment, trends in Q3 were consistent with last quarter. Clients across industries are navigating elevated levels of uncertainty around trade policy and resulting impacts to their businesses. We are also seeing clients carefully evaluate technology investments, which is resulting in a lower pace of discretionary spending in certain areas like products and resources. At the same time, cost pressures continue to spur demand for productivity-led and vendor consolidation opportunity across segments. And we see a growing pipeline of modernization projects that lay the foundation of AI-led transformation for our clients.

Now turning to segments. We delivered year-over-year organic growth in all segments in the third quarter. Financial Services led growth, driven by healthy discretionary spending trends in areas like digital engineering, legacy modernization and generative AI initiatives and improved spending among insurance customers, particularly in North America. Health Sciences was in line with our expectation and has remained resilient despite the uncertainty around government funding and trade policies. While we have seen pockets of discretionary spending pressure, it is being more than offset by the ramp of recent wins in payer and life sciences.

Products and Resources revenue growth has improved, and we are confident we can build off these levels in the quarters ahead as we expect new deal wins to ramp up more meaningfully in 2026. And Communication, Media and Technology grew organically and benefited from recent large deal wins that more than offset pockets of discretionary spending weakness in the quarter.

Geographically, North America once again led growth and was up nearly 8% year-over-year in constant currency, driven by our large deal success and Belcan. Outside of North America, demand trends in Europe and Rest of World remained stable, but not immune to impacts from recent tariff and geopolitical uncertainty.

Turning to bookings. On the trailing 12-month basis, bookings grew 5% and represented a book-to-bill of 1.3. After a strong performance of 18% year-over-year growth in Q2, we experienced some lumpiness in the third quarter and bookings declined by about 5% year-over-year. Our trailing 12-month annual contract value, or ACV, growth was consistent with TCV growth. Overall, our backlog remains healthy and our sustained large deal momentum provides us good visibility as we exit 2025.

Moving on to margins. Third quarter operating margin of 16% increased by 70 basis points year-over-year, benefiting from NextGen program savings and the Indian rupee depreciation. Utilization held steady at 85% for third consecutive quarter, up from 84% a year ago. These improvements were partially offset by the ramp of large deals and the dilutive impact from Belcan. Voluntary attrition remained low at 14.5%, down 70 basis points sequentially, the third consecutive quarter of sequential decline and down 10 basis points year-over-year.

A brief comment on H1B visas. Over the last several years, Cognizant has significantly reduced the dependency on visas while increasing local hiring and our nearshore capacity. We also stepped up our investments in automation and AI productivity tooling. We, therefore, do not expect a material impact to our operations or financial performance in near term as a result of the recent policy changes in the U.S.

Now 2 additional details. During the quarter, we recorded a onetime noncash income tax expense of $390 million or $0.80 per share. As we discussed last quarter, this charge is related to a deferred income tax asset on the balance sheet that is not expected to be realized due to the enactment of the July U.S. budget bill. Adjusted EPS, which excludes this impact, was $1.39, up 11% year-over-year. DSO of 82 days declined 1 day sequentially and increased 1 day year-over-year.

Third quarter free cash flow was $1.2 billion and represented 170% of adjusted net income. This compares to free cash flow of $791 million a year ago. As a reminder, cash income taxes in the third quarter were approximately $150 million, lower compared to our projections prior to the passing of the July U.S. budget bill. For the full year, we expect that reduction to be $200 million. Through the first 9 months of 2025, free cash flow is $1.9 billion and represented approximately 100% of adjusted net income.

During the third quarter, we returned $600 million of capital to shareholders through share repurchases and dividends, bringing the year-to-date total to approximately $1.5 billion. We are on track with our plan to return $2 billion to shareholders in 2025. This will bring total capital returned to shareholders since 2022 to nearly $5 billion. We ended the quarter with cash and short-term investments of $2.4 billion or net cash of $1.8 billion.

Finally, our M&A pipeline remains active, and we have ample flexibility to invest strategically in the quarters ahead while continuing to return substantial capital to shareholders. Now turning to our forward guidance. For the fourth quarter, we expect revenue to grow 2.5% to 3.5% year-over-year in constant currency, which is all organic. We, therefore, now expect full year revenue to grow 6% to 6.3% in constant currency, above our prior guidance range of 4% to 6%.

We continue to expect full year inorganic contribution of approximately 250 basis points. We are increasing our adjusted operating margin guidance to approximately 15.7%, which is the upper end of our prior guidance and represents 40 basis points of expansion. We continue to expect margin performance will be driven by cost discipline and SG&A leverage. This year, the fourth quarter will include the impact from a merit cycle compared to its Q3 timing last year. This will be partially offset by year-end seasonal margin strength.

We continue to expect free cash flow conversion to be approximately 100% of adjusted net income. This includes the benefit from lower cash taxes as a result of the U.S. budget bill discussed earlier. We expect our adjusted tax rate, which excludes the onetime tax charge, to be in 24% to 25% range. Based on our current visibility, we now expect full year tax rate to be closer to the midpoint versus the lower end that we indicated last quarter.

We are increasing our EPS guidance to $5.22 to $5.26 compared to our prior range of $5.08 to $5.22. This represents 10% to 11% year-over-year growth. Our expected weighted average diluted share count is unchanged at approximately 489 million. In closing, we are very proud that our guidance puts us on track to meet or exceed the high end of the initial guidance range we provided back in February despite a dynamically changing market compared to the beginning of the year.

While we are not commenting on financial expectations for 2026, we feel well positioned to carry this momentum as we look ahead and remain committed to the long-term financial framework we provided at the Investor Day earlier this year. With that, we will open the call for your questions.

[Operator Instructions] Our first question comes from Jim Schneider with Goldman Sachs.

Ravi, I wonder if you could speak to the new business pipeline you're seeing for smaller deals at this stage and whether you're seeing any kind of significant uptick there or not? And then relative to larger deals, are you seeing any pull-in or extension in terms of the commencement date for those large deal new bookings?

Thank you, Jim, for that question. Look, large deals have nicely balanced between the 2 swim lanes I've been talking about. Early on in 2024 and early 2025, a lot of it was consolidation productivity led. Now we are seeing a new swim lane evolve, which is AI innovation led, which is primarily agentic cycles, deploying AI into enterprise landscapes. And therefore, if you've noticed, our digital engineering business has grown at 8% in the last few quarters. Our infrastructure-led AI has grown by 10% in the last few quarters.

And our BPO business is rocking. It's actually growing at 10% again. And that we are starting to see. So it's a combination of productivity-led, innovation-led. I mean, I've always been saying that this is a double engine transformation. While you can apply it on software cycles, get productivity and transfer the lower cost of deployment for higher spend of software.

On the other end, you can apply it on -- you can apply agentic capital on enterprises. There's so much of infrastructure spend, which has happened. It has to create a build opportunity. And that's why I keep saying we are an AI builder. So we are seeing discretionary small projects starting to come back in financial services and health care. And that's all related to AI-led spend. So as you save on one side from software cycles, you transfer that money to innovation. So very healthy pipeline.

I'm excited about the large deals. I'm excited about the discretionary coming back on small deals, which is AI led. I mean so much infrastructure has been spent that it has to trickle down to services. And there's always been a lag between when hardware was spent, then the software and then the services. That cycle has shrunk now, and we'll be breaking that cycle. So the services spend is going to catch up because of the extraordinary spend on compute and AI infrastructure.

Our next question comes from Tien-Tsin Huang with JPMorgan.

Nice results. Well done. I want to ask on the revenue per employee. It looked like up 8%, operating income also better than that, up 10%. So just understanding the lift there and if it's sustainable or even structural given some of the AI returns that you talked about.

Thank you so much for that question. In fact, just a follow-up on the previous question. We're also seeing mega deals. Last quarter, we did 2. The quarter before, we did 1. So mega deals are also starting to line up because that savings can be underwritten for innovation.

Now coming to your question, this is an interesting lead indicator, revenue per person and margin per person. Revenue per person went up by 8%, margin per person went up by 10%. It's indicative of how we are becoming an AI builder company with platforms, intellectual property, software and services all bundled together. So we're excited about that. It's a combination of things.

Our fixed price managed services business is going up. It has gone up from 43% in 2024 to right now almost close to 47%. That gives us a chance to deliver work for outcomes and therefore, create more revenue per person and margin per person. It is actually going to transition, and Jatin has been talking about it, that we are a fixed price time and material and a transaction-based business. We are going to go from more fixed price, more outcome-based, more transaction-based, less time and material in the future.

So productivity has gone up 30%, which means there is more throughput you can actually create more throughput, share the savings, lower cost of deployment with our clients. So effectively, putting all this together, this is a very good proxy for AI services. And that's why we thought -- we've been tracking this, but we thought we should let analysts and investors know about it.

Yes. That's good proxy, good data point for us to have. Just on the -- my follow-up, then I'll ask on gross margin probably like I usually do. Just thinking about near-term gross margin performance potentially given the expected deal ramps and the mega deals and what have you. Any specific callouts on gross margin in the next couple of quarters?

Sure. Thank you for that question, Tien-Tsin. I would start by saying how we have executed for first 9 months, while you see the headline number a little soft, but on gross margin, we have been able to largely maintain the gross margins on an organic basis. The reduction that you see on a year-over-year basis is coming through on account of the consolidation of Belcan, which was expected when we did the deal.

So overall, we are quite happy that despite the ramp-up of large deals and investments that we are making, we are able to maintain gross margin in a very narrow range of last year in an organic basis. Going forward also, our endeavor would be to continue to look at 3 or 4 operational levers. And the top of that is AI-led productivity that Ravi spoke about.

The second is pyramid. You know we have invested 15,000 to 20,000 in recent college graduates and it's more than most. And it's significantly higher than last year. So we continue to improve the pyramid. And third is utilization, which you can see we have kept it at 85% now third quarter in a row. It is higher by 1 percentage point compared to quarter 3 of last year. So we feel we are making good progress on gross margin, and hopefully, that will continue to reflect in the numbers.

Our next question comes from Maggie Nolan with William Blair.

Can you shed some insight on how you're tracking the success of upskilling your employees with those AI-related skill sets?

Thank you for that question. I mean we are pioneering this effort. Early on, we were the first company and probably the only company which -- in our peer group, which speaks about percentage of code and software development cycles assisted by machines. That's at 30%, and we are constantly tracking to stay ahead of the curve. We are the #1 company on GitHub Copilot. In fact, we are the GitHub Copilot AI Partner of the Year. We have been the launch partner for Gemini -- Google Gemini Enterprise. We just signed a deal with Anthropic on cloud.

We have created a hustle inside the company that the only way you should write and the only way you should be assisted in software development is through machines. And that has become the way of doing work at Cognizant. In fact, we are on the Guinness book of records for the highest number of people on an hackathon concurrently. In fact, we ran that to create culture and create a permanency in the way we write software. We are the only company which is actually saying we are going to hire more school graduates than ever before. We are doubling those numbers from last year because we think we can actually create a significant productivity leap with our extraordinary training infrastructure.

So all of this put together, we seem to be on a pioneering opportunity. This is on one swim lane, which is software development. Of course, there is a ton of work on the agentic development, which is much more surface area, much more spend, more expensive. So our double engine, both on productivity and on innovation is deeply embedded with skilling, reskilling, hiring from schools, building productivity alongside machines, and that is the only way we want to create throughput.

And if we do this well, software has elasticity to be spent more. If we do this well, that money is going to be transitioned to agentic capital and working alongside agents for human workforce, I think, is going to be an amplifying potential of humans. So we have pretty much trained almost all our employees, more than 250,000-plus employees on AI-led skills.

That's helpful. And then should we expect large deal and mega deal signings to impact the quarterly cadence of revenue and margins in 2026? Can you help us think about the ramp over the course of the year from a modeling perspective?

That's a great question. In fact, if you notice, in 2023, our trailing 12 months range of bookings was in the range of $24 billion. And it's now actually at $27-plus billion. So we've had tailwind from '24 into '25. Our annual contract value is very nicely stacking up to the total contract value. In fact, our total contract value from large and mega deals has gone up by 40%, while the number of deals is 16 and so far, and we have another quarter to go. Just the TCV value has just significantly gone up. It's gone up by 40%.

So we think we have tail velocity going into quarter 4 as well as going into the next year on large and mega deals. I don't see any shift on that. And on the contrary, I actually see that on 2 swim lanes. In '23 and '24, the swim lane was productivity.

And in '25, we are starting to see we are starting to see innovation-led, agentic capital-led, much more expansive. I've always been saying one swim lane is software, another swim lane is Agentic. The Agentic is more expansive, more elastic and more immersive. We are seeing large deals on it. And in '23 and '24, a lot of it was Americas based. Now we are seeing Europe and Asia Pacific starting to be a part of it, and we are excited about the momentum we have created in Europe on large deals.

We'll go next to Surinder Thind with Jefferies.

Ravi, can you maybe talk about the partnership strategy here and how important it is to maybe partner with each of the major providers versus maybe being a bit more selective and becoming more of a partner of choice with maybe some of the individual providers, whether it's GCS versus Anthropic or OpenAI or however you're thinking about that strategy?

Surinder, thank you for that question. Partnerships traditionally were SaaS companies and classical software companies, of course, cloud-based hyperscalers as well. Now I would add more things to the mix. I mean, look, SaaS companies and classical software companies will transition the business logic to the Agentic layer, which they build on it. I mentioned this in my remarks that the machine was always with the software companies, and we were a system integrator.

Now we are AI builder, which means we have intellectual property platforms built. It's a very heterogeneous and a fragmented AI market. Our clients are not saying, come in with your capability. They're saying, come in with your machine. which means you have to actually have the platforms. It could be partner-led, it could be our own. And we are actually, therefore, investing into platforms and intellectual property.

In addition to that, we have this new thing because now the machine actually belongs to the frontier model companies, which is OpenAI, Anthropic kind of firms. In fact, that's one of the reasons why we partnered with Anthropic. So we are activating multiple swim lanes, our own custom platforms built on enterprise software companies where we have long-term partnerships, SaaS companies, but we also are partnering with frontier model companies because we could create custom AI agentic capital directly. And the Anthropic partnership is an indication of the particular swim lane. So it's a much broader partnership lens, including start-ups.

I mean I work with Writer, I work with Work Fabric AI. These are layers of value on top of the LLM. And some of those layers are owned by us, built by us, some of them are partnered. And of course, the frontier model companies allow us to create a swim lane on -- with the engine actually belonging to them, but we build the layers of service around it. So we think it's more expansive and more broad-based.

That's helpful. And then as a follow-up, can you maybe talk about the IP that you're building? And more specifically, you mentioned having upwards of 1,500 agents in production. How does that impact the revenue model at this point? Are you able to charge for some of that? Do you keep some of that IP? Or is it more of a core base and then you kind of custom build agents then.

I think it's a combination, Surinder. It's a combination. Look, on Flowsource, which is a platform which sits on top of Code Assist platforms, it gives us the opportunity to get better productivity and that productivity passes on to our revenue per person and margin per person metrics. Our other IP and platforms we are building are the ability to take the raw power of AI and make it enterprise grade, which means it could be the accuracy of the models.

Yesterday, we got a patent on changing the -- on a new way of pretraining a model, not based on reinforcement learning, but based on evolution strategies, which our labs got in. So we are building a platform around it. We have a platform around multi-agent systems, which means you could have agents talking to each other. One example is in TriZetto, our TriZetto agents talk to Salesforce agents and Genesys and the ServiceNow agents, and they actually deliver outcomes like ID cards and status of a claim automatically and auto adjudication of claims and stuff like that.

So the platforms are all about taking the raw power of AI and making it enterprise grade. It could be on accuracy, on responsible AI. It could be on new ways of pretraining the model, a variety of things which are needed to make it enterprise grade. There is so much infrastructure spend, which has happened. That value has to trickle down. And the use cases, which are now coming out, production grade, we are able to generate more of it because of the intellectual property we have built.

We are also closely monitoring partnerships. I mean the context engineering piece is a unique pioneering opportunity for us. This is a contextual computing era, which means you need to feed the context, could be the tribal knowledge, the workflows, the data flows, the hassle of a company and you have to feed it into the LLM and create a contextual agent who is much more productive than a generic agent. That actually needs -- it's a science which is evolving. So we're building intellectual property along with a partner of ours.

So I think this is going to be a platforms plus capability kind of a model. And therefore, I call myself an AI builder company, and we are pioneering that effort of transitioning from just a capability firm to a platforms plus capability. And historically, we had that culture with our health care business where a lot of it is platform plus services with TriZetto.

And our next question comes from Darrin Peller with Wolfe Research.

Just a financial question. Just when I look at the guide of 2.5% to 3.5% constant currency for fourth quarter, just what are the puts and takes there? Any early insights into how budgets are shaping up also into '26 would be helpful.

So as you can imagine, difficult to talk about '26 at this juncture. We will come back in January and speak about it. But overall, there is no major change in the demand environment. We continue to win share, and that is reflected in the superior execution of the quarter that went by. Quarter 4 seems to be a customary quarter 4 with its lower number of bill days and furlough. Nothing out of ordinary. Our guidance range reflects essentially, if things could go a little worse, then it's the bottom end. And if we can get some additional momentum in revenue and bookings, then it's the upper end. So that's how we have worked through quarter 4, and that's the full year guidance.

Just one quick addition there. I would say, look, the activation of AI-led innovation use cases, we've gone from 2,500 to 3,500 this quarter. So literally 40% jump. So the money you save on the software cycles on productivity is going to be underwritten to innovation cycles. That is triggering off very well. And I don't think CIOs are saying they're going to cut their budgets. Nobody has told me that. They're all going to -- they're all saying, how can you give me more value? And that's why we are benefiting out of this.

Okay. That's helpful. And then maybe just one quick follow-up would be if you could just discuss -- just you're obviously doing well with larger deals. Maybe just discuss the competitive dynamics for some of the large deals you're seeing and what's allowing you guys to continue winning them? And then how important is price in the discussion and maybe build into that what AI can do for you on pricing, if you could pass through some of your savings into?

Yes. Look, price was always a linear thing in the past because it was labor related and productivity of tooling was not in the mix. I would say it was a minority. Right now, pricing is productivity led, and it depends on how much you can use your platforms and how much you can use your AI tooling and the culture we have established now. So pricing is kind of linked to how fast we can keep that runway on productivity. Large deals on consolidation and productivity will always be price sensitive because they're done for savings and creating more velocity.

The innovation side of it, I mean, that's going to be much more -- that's going to be less sensitive to price because you are actually delivering new products and new services using AI. I don't see much of a change in the pricing. I would actually say if the other swim lane gets activated, which is innovation-led, you will get strength behind the pricing.

[Operator Instructions] And our next question comes from Yogesh Aggarwal with HSBC Bank.

Just have a question actually totally disconnected to the quarter and demand, et cetera. Just in the past few quarters, Cognizant performance has consistently improved and now you're growing almost at the top end of the peer range. But I'm sure you would have noticed as well the stock still is at a significant discount to the peer group. So I was just curious, any thoughts on secondary listing in India? I mean, is it something on the table and any puts and takes for the same? Just curious to know your thoughts, please.

Yes. So Yogesh, thank you. That is an interesting question. Cognizant's Board and management team regularly assess opportunities to enhance the shareholder value. Towards this end, we have been assessing a potential primary offering and a secondary listing in India with our legal and financial advisers. As part of this comprehensive review, which is still in its early phase, we are engaging various stakeholders from both India and U.S. to evaluate the implications of such a potential offering and listing. The process of a primary offering and a secondary listing in India by an overseas company is complex and involves multiple steps.

We view this as a long-term project. While no decision has been made and any offering and secondary listing would be subject to market and other factors, we continue to assess and review the idea and are committed to acting in the best interest of our shareholders. So that's our response, Yogesh.

Our next caller comes from Rod Bourgeois with DeepDive Equity Research.

So I want to talk for a second about financial services vertical. You mentioned improved spending there, and we are seeing some of that across the broader sector. Can you speak to what form that improved spending is taking in the financial services vertical? And in particular, are you now seeing those clients moving beyond AI for cost savings and into more AI-based reinvention at those clients?

Thank you, Rod, for that question. Absolutely. I think this is probably my fourth or fifth quarter where we have done year-over-year growth as well as from the start of the year, we have been sequentially growing in financial services. It's been one of our best-performing industry groups. I think the spend has gradually transitioned from cost takeout consolidation to more innovation.

I would say if you take the 3,500 projects we are doing on AI-led innovation, a significant chunk are actually moving from experimentation to enterprise-grade AI. And all the platforms I'm speaking about in the call, a lot of them are getting implemented in financial services. In fact, insurance, which is a part of BFSI has also started to spend. It's a sector which kind of was lagging a little bit, but it has started to spend as well. So we are very, very excited about the future of financial services.

Over the last few years, Cognizant per se, -- we have had quarters where we haven't performed in the previous years. But from -- the turnaround has started from the middle of, I would say, 2024. And here we are, it is actually one of -- it is actually our best-performing industry group. The spend cycles are great, and clients are actually innovating much more.

Every segment of financial services has accelerated in terms of spend and the discretionary is coming back because the value you get out of discretionary now is much higher because the cost of capital is high, but the deployment costs have gone down. So that is giving clients the confidence to experiment more and actually take it to production. So financial services will be one of our bellwether industries in 2026 as well.

Great. And then moving to health care. I mean, there's been some policy uncertainty in that vertical. You've also got the TriZetto asset. Just can you speak a little bit about the outlook for the health care vertical in general? And in particular, with TriZetto and all of the AI work that you're doing, are you seeing BPaaS as an increased opportunity there? Just any color on the health care outlook.

Absolutely, absolutely. In fact, if you look at health care, I mean, if you take the last 20 years, the number of surgeons and the number of doctors has pretty much remained flat. But if you look at administrative costs, they've probably got up like 600% to 700%, number of administrators in that business. So transitioning that spend to predictive care, I think, is the future. We have 200 million-plus members on our TriZetto platform. We own the BPaaS cycle.

In fact, you have -- you answered my question. BPaaS is our hottest offering. It gives us the opportunity to not just share our platforms, but equally, the operational strength of running health care operations, I think, is an important consideration. One of the reasons why our BPO business is 10-plus percent growth this year is also because of BS -- so we are very, very excited about our BPaaS offering, AI-led instrumentation in our TriZetto business and our lead in health care. I mean we are probably the #1 player in health care in the United States.

We'll go next to Jonathan Lee with Guggenheim Partners.

Good to see the outperformance here. You called out last quarter that you were expecting the 4Q exit rate to be just under 4% at the high end of the outlook range. Given the outperformance this quarter, can you help bridge the gap between the 3.5% at the high end of your 4Q outlook today and the 4% exit rate you pointed to last quarter?

Yes. I think this is 2.5% to 3.5% is the view that we have of quarter 4 as we look at next few -- next couple of months. It is -- we have great momentum in terms of winning the large deals. We have been able to execute better in quarter 3. And that will really decide -- I mean, continuing momentum on those factors will really decide where 2026 comes through. But overall, we are very happy with the way we have executed 2025.

As I spoke in my opening remarks, when we gave guidance since then, the environment has been very, very dynamic and still to be able to come in the last quarter and guide above the original guidance range is very heartening. So we have executed well, and we hope we'll continue to do so as we move forward.

Can you help us also better understand your pyramid initiatives and how you're balancing the needs of clients while managing margins, particularly as you move into higher-value AI-related work in vectors 2 and 3 that may require higher skilled talent beyond that of freshers?

So we have been very vocal about the fact that we see actually freshers and AI a very complementary strategy. And we believe that expanding pyramid at the bottom in our industry really helps us accelerate the organization's journey on AI. So from that context, we more than doubled this year, the number of freshers we took last year to this year, and that journey will continue one from a cost management standpoint of pyramid, but even greater context is how we can accelerate the enterprise to become more AI-ready and AI builder, as Ravi spoke about. And also, we are now hiring freshers in the markets, which is primarily our principal market being the U.S. So we are doubling down on this with a broader pyramid and a shorter path to expertise.

Thank you. And that does conclude our question-and-answer session. I would like to turn the floor back over to Ravi Kumar for closing comments.

Thank you so much for joining us today. We are very excited about our strategy of being an AI builder company. which is a combination of AI-led capability, platforms, intellectual property and partnerships, which allow us to be on those 2 swim lanes, one on productivity, one on AI-led innovation with a much expansive, elastic and a more immersive opportunity to serve our clients. So we're very, very excited about our future, and thank you again for listening to us today.

Thank you. This concludes today's Cognizant Technology Solutions Third Quarter 2025 Earnings Conference Call. You may now disconnect.