Hexaware Technologies Limited (HEXT) Earnings Call Transcript & Summary

Ladies and gentlemen, good day. Welcome to Hexaware Technologies Limited Conference Call for the Q2 CY '26 Earnings Call. We'll begin today's session with a presentation from the Hexaware management team, followed by a Q&A segment. I'll now hand the conference over to Mr. Niraj Khemka, Head of Investor Relations. Thank you. Over to you, Mr. Niraj.

Thank you. Hello, everyone. Welcome to Hexaware Technologies Q2 CY '26 Earnings Call. In the call today, we have with us Mr. K. Srikrishna, CEO; and Mr. Vikash Jain, CFO. In the course of this call, we will make certain statements which are forward-looking and may involve a number of risks and uncertainties. All forward-looking statements made herein are based on information presently available to the management, and the company does not undertake to update any forward-looking statements that may be made in the course of this call. In this regard, there's a full disclosure, which has been included in the investor presentation and the press release. We consider that as read. With this, I'll hand over the call to Keech. Over to you, Keech.

Thanks, Niraj. If you go to the next slide, please. The next one. So hi, everyone. Good to talk to you. At the highest level, I would say that I think the adoption of AI and where it works and where it doesn't work in -- especially in IT, I think it's settling down. And again, I believe that the level of compression that we have in the worst-case scenarios thought about, I don't think it's going to happen. Let me give you maybe two examples to bring out how this is settling, right? So just yesterday, we had this client for whom like five, six applications had to be rebuilt to build a portal, which went live in production in under 3 months from ideation. That's a great example of the use of AI. And in this case, the customer and the business is ready with what to do next once this program is done. But yet we have another larger enterprise customer where we finish what has to be done in 8 weeks in less than 4 weeks. But when we went back to the business for the next set of requirements, they said, hey, you will come back in 8 weeks. We are not ready. And we don't have -- we don't need this velocity. We are not quite ready, in fact, thinking about what we want next. So what this means is that adoption is not definitely about just technology. But it's not even about change management just in the tech teams, them coming up to speed and knowing how to deliver. But it is a larger change across enterprises, especially in the business side, which I think is going to slow things down in several scenarios. On the other hand, I think whenever Glasswing and Mythos gets out of stealth or even before that, when the lesser models, which are now available, people start discovering vulnerabilities at machine speed, I think it's going to unleash a massive amount of need for tech modernization. What we already know is that the vulnerabilities are coming from older systems. Open source, especially is a known area. And our belief is that it will become a critical imperative to fix everything that is old. So there's no way to manage the velocity and volume of vulnerabilities that get discovered and get organizations to a safe place without modernizing all their systems. So we think that will be a massive set of opportunities that is around the corner. During all this, we continue our pivot to AI every day, every week. So a few highlights. Now I'm going to keep this light because we are doing an AI Day in our Chennai campus on 21st. We had earlier planned for it to be shortly after today, but we just wanted like the cycle to get over and all of you to be fully free. So it's on 21st August in Chennai, and we will go into more details on several facets, but I wanted to give you a few highlights. We continue to launch new service per month. And that's one of the core principles we defined last July, that we will launch new service every month and also knock on 100 customers in 90 days after launch for every new service. Now we are well on target for the first one. We are getting close to meeting the second goal of 100 customers in 90 days for the second. The -- we're not just launching. We are scaling what we launch across all of our customers. We formed these champion quads for each business. In our case, the range is between 30 for the smallest unit to 75 AI champions for the largest units, who then are ensuring that we are executing proactively in all customers, the things that we are bringing to market. Zero License, which we launched earlier this year, is seeing outstanding traction in terms -- first, in terms of conversation. It's actually -- last quarter, I said AI and SDLC is the biggest driver of conversations and deals. That's still true for deals. But in conversation, actually, Zero License is rapidly getting to the same spot as AI and SDLC. We've closed the first few deals, and I'm going to give an example or two later today. But we have quite a large number in the pipeline. We've built by this time, and this number is going up on a weekly basis, 65 parcels in our platform for Zero License. What these parcels do is, for the 65 different SaaS platforms, we can discover the business logic underlying the SaaS platform. We published earlier in Q2 our first AI impact series. It's a book of case studies. There are dozens of case studies covering every vertical, every service line, showing the impact to clients. In fact, we will have our second book when you come and visit us in campus. I think over the last -- we've been talking about tokenomics, not using that word, but we've been talking about AI economics to our clients since April last year. It has suddenly become center stage for clients over the last 3 months. Not only is the cost of tokens being center stage, it is also, I think, in some ways, at the center of battle for IT budgets. How much will go to labor, how much will go to tokens. And I'm going to talk a little more on this as well. Now I think you would have seen last quarter, we announced or last -- late Q1, we announced that Shantanu is taking over additional role as U.S. hunting head, North America hunting head, apart from his role as H&I vertical head. A few weeks ago, we announced that Param Iyer and Vijay will respectively take over our Europe and Asia geographies. All three of them are in what I will call a Phase 2 transformation of our hunting team. Our Phase 1 happened over the last year or so, where we redid our hunting team to regain momentum, and that has paid results. But I think there is more work to do. So really three focus areas: large and proactive deal creation, significant expansion of team and extensive use of AI in the sales process. And that's again been there for a while, but really democratizing it, making sure every single salesperson across the world is able to use AI in every stage of the sales process effectively. So those are the three focus areas for transformation. We continue to add headcount. This is a third quarter of addition in IT, and our overall headcount addition was 708 for the quarter, about 179 in IT. Our attrition is very good, and our utilization rate has ticked up to what we think is a good spot for us to be in. We added a new center in GIFT City, and we added a new center in Bogotá in Colombia during the last quarter. On financials, I think we had a solid quarter in a difficult environment across revenue and profitability. You will hear more on it from Vikash. Our EPS, the underlying EPS growth is solid, but there are some hedge and translation losses that led to a compression in EPS. Those will recede shortly. And again, you will hear more details from Vikash. And as always, we have a solid cash position. But the highlight for us this quarter on our numbers was something at the back. We've had, I don't know, 3 years long ERP program. We cut over finally and fully. We've been going live with modules, but all modules, the last one being our financial modules, went live in the course of the quarter that just ended. And this is the first set of reporting that we're doing all from the new system, completely without a hitch. Next slide, please. I said I'm going to talk a little bit about tokenomics, right? So what we're doing with -- is experimenting with a lot of different models, and I've hidden out some here because of confidentiality reasons. We are experimenting with different models with two objectives. One, how do we maximize value for clients? We don't want to be on the side where, hey, we'll keep pushing token usage up for you and yet find a way to participate in the token economics. With those dual set of objectives, we are experimenting with 8 or 10 different pricing models for customers, input, which is the majority of the traditional model, output, but also outcomes, which are different, right? Output is, hey, I will build something for a fixed price, but outcome is that will drive x amount of revenue. So we are experimenting with all of them. And I'll give you an example of -- actually, before that, the key difference in approach, the most fundamental difference in approach is what you see in the two columns here, right? The historic, our proposals included human cost. Occasionally, they'll include the licenses, the hardware, the cloud usage, occasionally. The big shift now is that our proposals are also providing options for including token costs, but in a way that brings value to clients. Let me give you an example of the value. Somebody is using Claude, we built a harness in Claude that has nine different ways in which token consumption can be optimized within Claude by using different models of Anthropic. This is an early stage, but I think what this will -- the impact for us will be twofold. First, it will allow us to participate in what is going to be a significant portion of clients' budgets, which is tokens. Second, it will improve our profitability over time. So those are the two impacts, but we are in early days of experimentation with these models. Next slide, please. The second topic that I want to talk about in AI today is partnerships. I wanted to highlight the philosophy more than the specifics today. The philosophy is that we believe at some time, and that is kind of happening even now, that the functionality and capability of LLMs will converge. So the differentiation is not going to come from there, for our customers or for us. The difference will come from what sits above them, the vertical depth and the client-specific context. So we are adding -- we are forming three layers of partnerships. The first, the bottom layer, is the foundational example, Anthropic LLM layer. But the differentiation is going to come from the layers above it, domain models, whether it's SLMs or fine-tuned LLMs, which bring the context of -- which bring the domain context semantic into an organization processes and IT. The layer above is it's not just about the domain, but we also need to have a deep understanding of how you work on a daily basis. And that's the third layer of partnerships. And all of this will still be underneath the platforms and what we take to market, which historically has been service line focused, Tensai for ITO, RapidX for SDLC, and Amaze for data transformation. But what you will also see from us very quickly is an increasing range of vertical products and platforms. And in the platforms, horizontal platforms, I forgot to mention Agentverse, which we launched in the course of this quarter. Next slide, please. We continue to have solid momentum in deals. As always, there's a mix of very core transformation work, outsourcing and consolidation. So let me pick a few examples here. The first one is, I would say, a storied German biotech firm, right? In the first phase of -- this is a 2-year program which we're doing for them. In the first phase, we are consolidating all the vendors that are doing clinical data right now. Shortly after, we are going to start transforming how the clinical data environment is built and convert that to an agentic platform. Second example is a large capital markets firm. They have a product which they feed securities to and they get cash flows. Now what is happening is that for years, data is given, but all they get is a PDF as an output. There's nothing more they can do with the data they have given to this platform. This is, I think, a key archetype for Zero License, okay? You're giving your data and renting intelligence back from the platform. So we are -- we have a program, which will happen in just a few months to exit them completely from this vertical platform into a modern core that we're building using AI and agents. Another large capital markets firm, AI-led middle office transformation in cash and settlements. I'm going to talk about a couple of modernization deals. So one of the things, and I will talk a little more about this, is that scale modernization deals are more and more in our pipeline and deals. So this is a client which is an interesting business model. You go to a restaurant, you hear music, somebody gets paid a bit of money for it. And that whole thing is managed by an organization. In this client, 2 years ago, we had lost an outsourcing deal because the client was very wary of handing over a legacy system from the incumbent. What we're doing now is to completely exit them out of this legacy system. And this is a north of $10 million deal, double-digit modernization deal. I'll just stick to the modernization theme. If you look right below there, leading CRO, CRM transformation, again, double-digit million modernization deal. And then we have three outsourcing deals, but one of them also has some other interesting things. So let me speak about each of them. A large university based in London, there are three parts to the program. There's outsourcing, all their tech work and BPO work. There's transformation. But the third part is the most interesting, which is that universities in many parts of the world are under stress because of political issues, enrollment from foreign students is going down, which puts a stress on their numbers and their enrollment and fees. And we are creating a shared service tech and BPO organization for the entire industry, which will be led -- I mean it will be a joint effort between this one leading university and Hexaware. The second one is -- and it's a reflection of our increasing momentum in APAC, large financial services firm in ANZ, Digital ITO. The third is, again, in APAC, a fintech payments firm where we're doing outsourcing and transformation of their entire payment systems. So that's some of the deals that closed in the quarter. With that, I'm going to hand over to Vikash.

Thanks, Keech. Can we move to the next page, please? As Keech mentioned, we are now live on the new ERP system, and this was our first quarter of financial reporting from the new platform. Now this transition required significant cross-functional effort, and I want to acknowledge the strong execution from various teams that enabled us to complete the quarter close and reporting cycle successfully. In terms of the financial highlights, revenue for the quarter was $405 million, reflecting a 4.4 percentage sequential growth. The sequential growth was primarily volume led, supported by a higher number of billable days during the quarter. In dollar terms, approximately $9 million of the sequential growth came from volume and around $5 million came from calendar benefit. Volume growth was supported by net headcount additions that you heard Keech speak about, the full quarter impact of the hires we made in the previous quarter and an improved utilization. On margins, reported EBIT for the quarter is 13.6 percentage, up about 60 bps sequentially. If you look into the various puts and takes, the sequential EBIT movement was driven by 160 bps of tailwind, which was from a combination of FX and calendar. We had an operational improvement of close to 30 bps, primarily driven by utilization. And these tailwinds were partially offset by Q2 seasonality and investments what we made of close to 70 bps and higher people investments of close to 50 bps. Now when I speak about the seasonal expenses, those are primarily our annual client event that we do. And I've also clubbed one of the M&A-related diligence expenses as part of this and higher CSR. Moving on to the next one. Some color on the unit level performance. Growth was broad-based. Five out of the seven verticals delivered both sequential and year-on-year growth. Year-on-year growth was led by health care and insurance, M&C and banking. Sequential growth was led by health care and insurance, M&C, PS and banking. Sequential softness in T&T reflects a tighter external environment, particularly in the Middle East. If you go a bit on the unit level, financial services continued to deliver both sequential and year-on-year growth. Revenue from the GSE client remains stable, and we expect healthy volume growth in financial services over the coming quarters. Healthcare and insurance, strong sequential and year-on-year growth, driven by the large deal ramp-ups and broad-based growth across Europe. In our prior earnings call, we had indicated that H&I would grow faster than the company average, and this has been playing out consistently quarter after quarter. M&C, deal closures are now converting into revenue, with growth coming from both existing accounts and new logos. Now M&C, if you recall, early part of last year, because of tariffs and everything, was facing a lot of headwinds, and we started seeing some green shoots in terms of the demand and getting converted into revenue. M&C has delivered year-on-year growth since H2 of last year. We expect M&C to be a full growth -- full year growth contributor. Professional services, what you see on a sequential basis, it reflects a bit of a seasonality. The largest client in this vertical follows a July to June calendar, and the year-end cycle typically drives higher budget utilization and revenue tailwinds. Banking delivered strong sequential and year-on-year growth. The vertical has shown consistent momentum in the last few quarters, and we expect it to remain strong throughout CY '26. Travel and transportation, as I spoke about earlier, remain impacted by macro conditions, particularly in the Middle East. On geos, the good news is every single geography delivered sequential and year-on-year growth. Europe returned to strong growth and is expected to lead full year growth driven by account ramp-ups and new logos. APAC delivered a strong sequential and year-on-year growth despite a challenging external environment in the Middle East. It's driven by the other subgeographies which continue to perform well. More commentary on the full year outlook will be provided by Keech later during the call. Let's move to the next page. We continue to add meaningful clients to our client base. And one of the ways we track the broad-based growth is by looking at number of clients, which are contributing more than $10 million in annual revenue. We now have 34 clients contributing more than $10 million annual revenues, an increase of three clients on a year-on-year basis. Let's move to the next one. This slide summarizes key operating metrics for the quarter. In terms of our revenue mix, on-site mix moved up slightly during the quarter, driven by rebadge deals. On a year-on-year basis, the offshore mix, if you look into it, has improved very meaningfully. Headcount during the quarter, the net headcount addition was 708, with close to 180 additions in IT and close to 530 additions in BPS. This marks the 12th consecutive quarter of IT headcount addition. BPS headcount addition is in anticipation of seasonal volumes increase in H2, with revenue realization expected in H2. Attrition remained stable at around 11%. Utilization closed at 84.8% up 220 bps sequentially. Now starting this quarter, we have made change in the utilization reporting methodology to exclude employees working on platforms. This improved utilization by about 80 bps and prior periods have not been restated. However, even excluding this change, the utilization remained very healthy, upwards of 84%. We expect utilization to remain broadly range bound in the 83 to 84 percentage range going forward. Let's move to the next page. Closing cash balance, Keech spoke about, was close to $176 million. Now as you are aware, during this quarter, we went with ERP transition. Now in the period of ERP transition, as would be expected, there is a blackout period when there are data transitions taking place from one system to the other. That did have an impact at the early part of the quarter in terms of the invoicing. However, the account teams worked proactively with the clients to support the cash conversion. Quarter-on-quarter, you see a cash balance reduction of $45 million, but that's after making close to $105 million of payout, $55 million in dividend, $27 million towards acquisitions and close to $24 million towards incentives pertaining to CY '25. The DSO for the quarter is a bit elevated, but it's in line with what we had expected, given the ERP transition. We expect DSO to normalize to the 70 to 75 days range by the year-end. Cash conversion, despite all of this, remains very strong and our LTM OCF to PAT was close to 125 percentage. ETR for the quarter was at 25.1%. We reiterate our full year ETR expectations to be between 25% to 26%. On EPS, we delivered strong operational performance for the quarter and EBIT margin expanded by 60 bps. However, this was not fully reflected in EPS because the quarter had material hedge and translation losses of approximately $8 million. Now these were driven by the forwards which were taken in line with our hedge policy. One call out from a future perspective. Based on the current hedge book and the June end exchange rates, we expect these losses to reduce over the next two quarters, with approximately $5 million in the next quarter and $3 million from a Q4 perspective. Those were the financial highlights. With that, I'll hand it over back to Keech. Let's go to the next page.

Thank you, Vikash. We feel very good. I feel very good about where we are as a business, and we will continue growth momentum. But yet we are reducing our guidance for the calendar year, primarily because of a narrow runway left. But I'll talk more on that. Deals, like I said, we feel very good about where we are, outstanding pipeline, outstanding wins and continued every month. The few -- I've spoken about some of the deals that closed, but I'll speak about some of the patterns here. One, we are seeing increased momentum in legacy modernization deals greater than $10 million. In the past, I spoke, we've got lots of deals in the $1 million to $2 million kind of -- and some of these will get to larger ones. And we have begun to see both in closures and pipeline, that we're seeing more deals in the greater than $10 million legacy modernization. We won a consolidation deal in another top 15 client. We are now down to three strategic vendors in this client. There is a new category of deals that have begun to emerge. There is, I would say, a handful of RFPs that started in Q2. These are very quick -- the ones that we've seen so far are quick decision cycles. We've already won two. I expect this pattern to go up. The pattern is that customers are doing an RFP to pick a strategic AI partner. This is decoupled from people that may be doing other work. And the first one we won, for example, they said, hey, here are kind of three use cases or four use cases, basis which we will evaluate the capabilities. But whoever we pick basis, this is essentially going to be our strategic AI partner. And in this example, we won it. The first four use cases, I think will get delivered in 2 months. And these are small currently, but we expect that we will spread our wings substantially across the enterprise, not just in IT, but also in the business as -- and we think there'll be more archetypes of these deals. I gave one example of a win in Zero License and there are -- and one archetype. There are three archetypes that we are currently seeing with traction. I won't talk to three archetypes, but we're seeing three specific archetypes where there's solid traction. Many of these deals are currently small. But again, we expect that most enterprises are picking one or two to prove the value, but we think they will scale across the enterprise as we deliver on proof of value for the current cases. Now revenue outlook, like I said, we feel very good about where we are on deals, but our pathway to 7.6% is narrowed. Specifically, I will say two things. One, deals that we won earlier in the year, there is a ramp delay, okay? In some cases, due to the Middle East situation, other reasons in other cases, but there's a delay in the ramp. Now much of that ramp-up is going to happen late Q3 through Q4 and of course, a worsening macro. You saw our [ GTD ] numbers, right? It is materially worse than what we thought it will be. And that's the place where potentially macro has the biggest impact. Given this, we are reducing our guidance to a range of 6% to 7%, including 50 bps from the CP rebadging deal that we announced earlier. Quick point about that, I mean, we announced it as an M&A deal. That is how the deal was done. However, it was done in that way for tax structuring reasons. It's a single client where we took over a contract and rebadge the people associated with that contract or set of contracts and took over the people associated with that set of contracts. The new guidance implies at midpoint a 2.7% CQGR from this point, which we're very confident of delivering. You heard detailed commentary on verticals from Vikash. But just to summarize, and this is actually the same set we put even last time, H&I, banking, and M&C are expected to lead growth. The only change is H&I was second on the list. We will now put it on first in the list. PS and FS will follow, and travel will lag materially due to macro. On margin, we are reiterating the margin guidance of an EBIT margin of 13% to 14%. With that, we will pause and do questions.

[Operator Instructions] Our first question comes from Vibhor Singhal from Nuvama Equities.

This is Vibhor here from Nuvama. Keech, a couple of questions from my side. Interesting that you mentioned that what clients are interested today is an application layer, which sits on top of the various LLM models, which provides them the flexibility to switch from one model to the other model. There's also the angle of optimization that we've been doing. So in terms of those, just wanted to check, are these conversations more at the supervision level itself at this point of time? Or are we seeing these conversations kind of culminating in some deals also? And when do you think these could actually translate into real large deals, the kind of transition that we probably saw in 2018 when we had similar deals which started coming through for the Claude adoption, which had multiple facets of technology that we were taking care for the client. So do you think there's a similarity between what happened then and what is happening now? And when could we see anything, let's say, material, which could basically give growth or boost on that part? Any color on that would be really helpful. And then I'll have a follow-up.

So I'll say kind of two or three quick things, right? First is harness, which sits on top of scaffolding, that sits on top of model or models is the term that is for what you just said, right, picking the right models. We built -- as many companies, but even labs themselves have built harnesses. They may not necessarily be only for picking across providers, but they could also be for picking within, say, Anthropic has many models or OpenAI has many models. So you could use the harness to pick different models for different types of use cases. I don't know that this will translate in of itself to large deals by itself, okay? So I think deals could come from different patterns, but I don't see this particular harness leading to large deals in of itself.

Got it. Got it. Got it. And you also mentioned a part in which very interesting kind of deals have evolved in which the clients are asking you to basically develop a use case and then basically deciding on the basis of that. So what are the kind of selection parameters which the clients are looking at? Is it the usual that how many people do you have certified on Anthropic or other platforms, if those certifications are available as yet? Or is it how robust the use case solutions that you provided is amongst the service vendors? So how does the competitive intensity here or let's say, how does the competitive differentiation work here for us, service vendors?

So I will tell you what we are doing, okay? And I spoke about use of AI in our sales process. So this we've been doing for, I'll say, 1.5 years, maybe 2, but we are now kind of scaling it, okay? So our preferred model of selling is no presentations, no proposals for transformation programs. We built a POV or some small version of what the customer ultimately wants built through the selection process. And we show it to them. This is how our sales process works today. So it's not about how many certifications you have and whatever else. We are saying, hey, there are four use cases the client defined to pick. We build solutions for the four use cases and take it for our solution defense. And that's to show and tell of that. And of course, these demos use synthetic data. But we say them, hey, if you kind of give me real data and if I can integrate with XYZ systems of yours, we can get this in production pretty quickly. This is the model for how we are selling. And then, of course, right now, this way of selling is, I'll say, like with some percentage of the company, we want to make that with 100% of the company.

Got it. Got it. Just one last question from my side. On the guidance part, the narrowing basically part of the guidance, which we've done from, let's say, 7.5% earlier, too. Now the midpoint being at 6.5%. Is it mainly driven by some of the deals which we expected to ramp up and they have kind of -- maybe there's a delay in ramp-up of those deals? Or is it also driven by the fact that the overall macro has kind of either remained the same or worsened. And so some of the deals which we might have expected have also not come through?

So it's both. But the second macro, I would say, is largely restricted to travel and transportation. It's not that deals haven't come through. Customers have cut. Existing clients have cut in the airlines business, their spends, their budgets. So you saw that T&T like shrunk 18% year-on-year. The deals are, I would say, not even delayed decision-making, but there were some delayed ramps. There are three, four deals for us that we expected will start ramping in Q2, which are now starting to ramp in mid- to late Q3. So we don't have the runway left to make up for it. So we called hence, the narrowed guidance.

Our next question goes to Aditi Patil.

I have three questions. My first question is on -- so I wanted to understand how much of our business has already gone through a cycle of AI-led deflation during renewals? And could there be multiple cycles of AI compression as the technology advances? My second question is the delays which we are facing in deal ramp-ups, they are in which verticals? And my third question is on the deal sizes in the new AI-led demand areas such as Zero License and the new emerging areas, which you mentioned about strategic AI partner-related deals. So what are the deal sizes in this area?