Good evening, everyone and welcome to the first quarter earnings press conference of HCLTech. We are joined by our senior leadership, led by CVK, Shiv Walia, and Ram Sundararajan. CVK, over to you to start the proceedings.
Good evening, everyone. Thank you for joining us today. We began FY 2027 with a focus on growing our advanced AI-led offerings, increase our relevance with clients, and capitalize on the full range of AI-related market opportunities in pursuit of becoming the world's best AI solutions provider. Our intent is to benefit disproportionately from the AI-native and AI-amplified opportunities, which together represents the fastest-growing pool of enterprise spend. While in AI-disrupted services, we intend to innovate faster than the market to stay ahead of the deflationary curve, rather than be defined by it. The fruition of this is reflected in our growing advanced AI revenue. Our advanced AI revenue for the quarter stood at $171 million, marking a 10.6% quarter-on-quarter increase and a 62.1% year-on-year increase. Our annualized advanced AI revenue now stands at $688 million. Let me delve into the business performance for the quarter.
AMJ is typically a weaker quarter for us due to planned revenue declines driven by productivity commitments in certain large managed services contracts. Despite the seasonal headwinds, we delivered a good all-round performance. Our revenue declined by 0.5% sequentially and increased by 2.6% on a year-on-year basis in constant currency. Our operating margin stood at 16.9%, an improvement of 40 basis points quarter-on-quarter and 56 basis points year-on-year. All growth rates referred to here are in constant currency. Our services business grew 3.5% year-on-year and declined by 0.7% sequentially. Our IT and Business Services grew 4.2% year-on-year and remained flat sequentially. Our Engineering and R&D Services grew 0.3% year-on-year and declined 3.7% sequentially. HCLSoftware annual recurring revenue now stands at $1.063 billion, a 2% increase year-on-year in constant currency. Revenue grew 2.2% sequentially, while declining 5.3% year-on-year. Recently, we've completed the acquisition of Jaspersoft.
This adds the visualization layer to our data management portfolio. What was missing was the visualized layer, the capability that takes governed, trusted data and delivers it as actionable insight to business users who need it. From bookings perspective, the quarter was very good. It was well-balanced across verticals, segments, and geographies. Our net new booking for the quarter was $2.4 billion. This is our highest ever Q1 booking till date. Of course, this $2.4 billion excludes the recent mega deal, which was signed in early July, and so not part of Q1 booking. We were selected by a Europe-headquartered Fortune Global 50 firm as the technology partner to accelerate AI-led transformation and management of their global digital workplace and enterprise networks. On the people front, we are 223,889 people as of 30th June 2026, which is a net decrease of 3,292 compared to the previous quarter.
Our attrition stands at 12.7% on LTM basis. Our AI strategy execution is showing up in a 3.3% year-on-year increase in revenue per employee, which has gone up every quarter for the last five quarters. With respect to our AI updates, as you know, our AI strategy is anchored on five strategic pillars. On each of these pillars, proactive transformation of our services, building differentiated IP, expanding AI-led services, scaling AI partnerships, and developing AI talent, our progress has been very impressive, and you would find several evidence of this in our investor release. I also want to spend a minute to comment on a recent strategic investment of $151 million in Sarvam, India's full-stack sovereign AI company. This investment reinforces HCLTech's position as an AI innovator, not just an adapter, underscoring our commitment to building, co-innovating, and shaping the next wave of AI.
By combining Sarvam's research depth with HCLTech's global enterprise relationships, our engineering expertise, and software IP, we are creating a differentiated full-stack AI platform spanning models, platforms, applications, and managed services. This meaningfully opens up the Indian sovereign AI market for us across industries and government sector. While the world's attention has been captured by the ever larger general purpose models, the real value for enterprises often lies in smaller, specialized models trained deeply on the language, data, and workflows of a single industry or a client. These models are faster, more cost efficient, and more accurate where it matters, and the window to establish leadership here is open now. I would also like to share an important strategic initiative of HCLTech: a full stack AI offering powered by AI data centers. Global data center demand is set to nearly triple by 2030, thanks to AI.
And in India, that growth is expected to happen at an even faster rate. Sovereign data requirements are increasingly mandating that workloads for governments, enterprises, and even global consumer platforms be delivered in-country. The business is shifting from physical infrastructure to higher value AI-ready full stack solutions; o ne we believe will be a new growth vector for HCLTech. We will combine our capabilities across AI data center design, DevOps, and cloud operations, as well as our software portfolio with a new AI data center business for this. In this regard, we will make a strategic investment up to INR 3,500 crore and with the potential to scale to 50 MW of capacity. Next topic is our pipeline and market trends. Pipeline remains healthy. Looking at the broader market, on one side, we are seeing strong, sustained growth in both AI-native and AI-amplified services.
And on the other side, where AI-disrupted services, which is the more traditional commoditized work, continues to be optimized further as AI-enabled automation takes hold. We are well-positioned to grow our business further through very clear strategies to address each of these three segments. Looking ahead, we are retaining our FY 2027 guidance of 1%-4% at a company level revenue growth and 1.5%-4.5% for services, both in constant currency. Margin guidance is also retained at 17.5%-18.5%. For clarity, this guidance is for organic growth and does not include acquisitions, including the recently concluded Jaspersoft acquisition. With this, I will now request Shiv to share more details on our financial performance.
Thank you, CVK. Good evening, everyone, and t hank you for joining our Q1 financial year 2027 earnings call. I will begin with an overview of our financial performance for the quarter, starting with the revenue performance. Total revenue for the quarter is INR 34,579 crore, which is up 1.8% quarter-on-quarter and 13.9% year-on-year. Services revenue is at INR 31,748 crore, growing 1.7% quarter-on-quarter and 14.9% year-on-year. Software revenue was at INR 2,960 crore, which is up 3.6% quarter-on-quarter and 4.7% year-on-year. In terms of profitability, our EBIT came in at INR 5,831 crore, representing 16.9% of revenue. This reflects growth of 3.8% quarter-on-quarter and 18% year-on-year. Net income came in at INR 4,624 crore at 13.4% of revenue, up 3% quarter-on-quarter and 20.3% year-on-year.
In terms of update on return on invested capital, ROIC, o ur ROIC continues to improve, supported by our focus on profitability and efficient capital management. Company level ROIC for the last 12 months is at 40.7%, up 257 basis points year-on-year, and services ROIC is at 47.8%, up 260 basis points year-on-year. HCLSoftware ROIC is at 21.6%, up 75 basis points year-on-year. Now, an update on cash generation and balance sheet. Our cash generation remained healthy. Over the last 12 months, free cash flow was at INR 17,843 crore, and operating cash flow was at INR 19,258 crore. The balance sheet remained strong with gross cash at INR 27,058 crore and net cash as of INR 26,907 crore as of 30th of June 2026.
Operating cash flow to net income conversion was at 107%, while free cash flow to net income conversion was at 99% for the last 12 months. Our total DSO, including unbilled receivables, was at 86 days, an increase of two days quarter-on-quarter. Now, moving on to shareholder returns. Our normalized diluted EPS for the last 12 months was at INR 66.99, up 4.5% quarter-on-quarter and 6.9% year-on-year. Including the one-time impact of New Labour Codes, diluted EPS was at INR 64.25. The board has declared an interim dividend of INR 12 per share for the quarter. The record date is 17th of July 2026, and the payment date shall be 27th of July 2026. This brings the last 12 months payout to INR 60 per share, effectively distributing 93.2% of our net income. That concludes my update.
I would now like to hand over to Ram for an update on HR matters. Over to you, Ram. Thank you.
Thank you. Good evening, everybody. CVK has already called out the closing headcount and attrition. I'll probably not try and repeat that. You need to look at the quarter-ending headcount in line with the revenues. You will see the productivity improvement. Revenue per employee has improved. We added 1,056 freshers this quarter. That, again, is in line with the plan for Q1, which, again, usually is a soft quarter, so everything gets reflected accordingly. Attrition, again, if you see, it's pretty stable now for a good 10 quarters. If you see here for the last 10 quarters, attrition is stable at 12.7%. I probably will stop there, and maybe, we'll open the floor for questions.
Anyone? Yeah. Yeah, please identify yourself.
Hi, good evening. This is Subhayan Chakraborty from The Economic Times. I would just move to the very interesting announcement that you made with regards to the INR 3,500 crore being moved to this new data center business. If you could just give a little bit of color as to what the financing aspects of that would be, and if you are partnering with some of the established players?
This is AI data center investment which we will make. Obviously, we are in conversation with several clients, and we are very close to establishing our first client with some committed capacity. We're starting with the initial investment of INR 3,500 crore, but our bigger vision is to grow it to a 50 MW. But the most important element here is, while it is AI data center business, we are playing fundamentally a different game. Normally, you look at this data center capacity in megawatts and gigawatts, and that becomes the product. Like megawatt and gigawatt becomes the product, and that's for infrastructure companies. For us, the megawatt is just the anchor. Our whole value is in delivering full stack AI services, which means it's data center, it's the GPUs, it's the models, it's the applications that we will deliver on top of it.
So, the overall value creation is significantly of a very different magnitude when you really look at this as a full stack, and that's really what we want to play. We are also looking at leveraging the data center that we are building for delivering managed services and outcome-based contracts for global clients, because t oday, token costs are a very important component of the overall delivery. So, we having our own capacity and our own models and solutions to deliver managed services and fixed price contracts is another very big advantage for global clients.
Hi, I'm Shashank from Outlook Business. HCLTech is one of the first IT services companies in India who have invested in a native foundational LLM company like Sarvam. So, my question is, how do you plan to monetize this investment? Are you planning to offer go-to general purpose model collaboratively with Sarvam, or you're planning to go towards the enterprise AI solutions?
I think, as I said, there are four significant opportunities, and first and foremost is the opportunities in India, both in the enterprises and in the governments, both state and central, and public sector undertakings. So, that's a big opportunity. Each of these large entities are looking at leveraging AI in a very meaningful way to really transform their businesses, especially in the financial services, insurance, and these kinds of verticals . While Sarvam will provide the models and the full stack capability, we will be the system integrators. So, we think there is a good opportunity there, both with the government and enterprises. And globally, today, the entire AI space is evolving with two broad principles. One is a zero-trust kind of a principle where no data leaves any enterprise for either data transfer, prompting, reasoning, the data remains within the enterprise, so there's zero- trust model.
It is also going to be a tiered model. Tier 1 will be a model which could be an SLM, which is within your enterprise, which can be used to train your own enterprise. Data can be used to train the model, which can do most of the work that you want to do within your enterprise, and only go to a frontier model on things which really need. That's a Tier 2. So, I think it's a hybrid tiered approach. That's what is going to happen in the global enterprises. Here, having a partnership with a foundational model firm will help us. So, we have enterprise relationships globally, and we have very deep engineering capability and lot of software products. This, along with Sarvam, will become an excellent solution for global markets. And some countries will also be looking at creating their own sovereign AI.
So, we can kind of do this together with Sarvam. These are the four opportunities we are looking at.
Good evening, gentlemen. I'm Shakshi Jain from Informist Media. One question for Mr. Vijayakumar. If you could offer some commentary on demand across segments as well as geographies, if there was any impact of the West Asia war seen in Q1, and how do you see things going forward? Also, you did mention the company aims to stay ahead of the AI deflationary curve. If you could elaborate on that? One question for Mr. Sundararajan as well. If you could give us some insight into the company's lateral hiring strategy for FY 2027? Last quarter, you mentioned fresher hiring numbers would be similar to FY 2026. Now, in this other segment, mid and senior talent pool, if you plan to exit FY 2027 with a net addition in headcount? Also, FDEs, what is the number currently? How do you plan to end FY 2027? One for Mr. Walia.
With the recent acquisitions of the company, right, can we expect an increase in the D&A to revenue ratio of the company? It's been around 3.5%, give or take a few basis points, so c an we expect an increase on that front going forward?
Okay. Let's address one by one. You talked about the demand environment. I think you should look at our booking. $2.4 billion this quarter is the highest-ever booking in Q1, so t hat's definitely a very positive indicator. And we continue to have a similar size of qualified pipeline as we had in the beginning of last quarter, so p ipeline continues to remain robust. Of course, there has been some impact due to West Asia. We did talk about it in the last quarter. It started in March, and some of that is continuing. Some of the discretionary spend softness continues to be there, but w e see a large pipeline, very healthy booking. We already announced a mega deal. We expect a strong booking even in Q2. So, I think overall, I feel quite positive about the overall outlook. Then, was it Ram?
Yeah. Your question was about the talent. So far, as our hiring plan is concerned, it's going to be in line with the revenue guidance. If we split that into what we do for fresher hiring versus lateral hiring, and I think you had a specific question about FDEs. Fresher hiring, as I said, it'll pretty much follow the trend that we had for last year. For a couple of years now, we have moved away from trying to commit to an annual number, and instead create a rolling quarterly plan, and that's what we will stay with. In aggregate, at the end of the year, this year would be similar to what we did last year. Our focus again is going to be not so much on numbers.
If I have to compromise on numbers, and if that results in increasing the percentage elite engineers that we will hire, that will be our focus. Right? The reason why we do that, it's exactly for the point that you made, which means, these elite engineers over the next two to three years should become part of the FDE cohort. That's the plan. As far as lateral hiring is concerned, it's always need-based. We don't go in with a fixed plan for a quarter or for that matter, even a month. It's going to be need-based. Again, it will just follow the revenue projection for each quarter, what capacity we have, what's the utilization, and what we need to build. That's the approach that we have always taken, and that'll continue. Specifically recruiting for FDEs, it'll be a mix of both build and buy. Right?
If we have to go for a talent buy, it's got to be more opportunistic for immediate needs. Otherwise, it's going to be on the back of the investments that we are making for building the talent internally. Right? We do have a target in place in terms of how we build FDEs internally. Every delivery unit will have FDEs built, and we are making investments for that. And we have been talking about the skilling investments for the last six quarters now, s o, all of that will add up to how we build FDEs.
Yeah. Your question regarding D&A going up. Yes, it will go up compared to what we have right now, because with any acquisition, a ll the amortization is going to be there in the P&L for initial couple of years. That's what we expect to happen.
When you say it will go up, by how many basis points?
I have not done the exact math in the sense we're still in the process of doing the purchase price allocation accounting for this. But I don't expect this to go up significantly.
All right. Okay. Thank you.
We have one more? Okay.
Hi. Just going back to the Sarvam investment for just one moment. I was just wondering, are you, sir, talking about offering Sarvam models to your clients as part of the new AI stack that you're talking about?
Yeah. I think it'll be very dependent on what the client requirements are, but that's a possibility.
Okay.
Now, we'll take questions. Okay. Yeah.
Hi, I'm Rishabh with NDTV Profit. Just a couple of questions. Telecom discretionary spending, the weakness was expected throughout the calendar year. Have spending trends improved in the first quarter? What have you seen on that front? Secondly, did the manufacturing and retail client, those budget cuts, did they continue in Q1? Do you have any visibility on that? Also, are procurement delays and slower decision making, are they still impacting deal closures? These three. Thanks.
Yeah. Sorry, what was the first?
Telecom.
Telecom, right, telecom. Telecom, as we had called out last time, there is certainly a reduction in the discretionary spending, and that continues. We don't see the situation improved in the last three months. Retail and consumer business, I think our retail and consumer business is doing very well. It's grown 10% year-on-year. This quarter, we have grown 5%. On the back of, we had announced a very large win two quarters ago with a large global retail apparel company, and that client has ramped up fully. We've executed that well, and that's causing the good growth in revenue. Manufacturing has some challenges due to different reasons, and some, like automotive and all, are under stress. So, that situation, I don't think has changed, e ven though we won some good engineering services opportunities in the automotive sector as well.
Thank you.
Now, we'll take questions from media who've joined us virtually.
Sorry?
People who've joined us virtually, we'll take questions from them. Over to you, Nitin.
Thank you, Ashutosh. For our friends who have joined us virtually today, please follow these instructions to ask a question. For people joining on the Webex mobile app, tap the three dots on the floating panel and select the Raise Hand icon. For people joining on the desktop app, please click the Raise Hand icon at the bottom right of the participant panel, and for people joining as audio-only participant, please press star three to raise your hand. Our first question, gentlemen, is from Jas from the Mint. Jas, please go ahead and ask a question.
Am I audible?
Yes, you are. Loud and clear. Please go ahead.
I just want to understand that when a year in the past, I think somewhere in October last year, made this foray into data centers. In an interview, you had mentioned that entering into the data center space was not on the cards, s o, what explains this change? Second, a related question, when you say that the GPU applications and models will be yours, that you all will be building layers on top of what you'll have to offer, c ould you elaborate a little on that? What some of the partnerships, how you all will be going about some of this work? And third question, chief, if I look at the revenue from the top five clients, that has declined 4% sequentially. So, this softness do you see, is it more than what you had previously baked in?
If you can also throw some light on the demand scenario in and around this, that would be helpful. Thank you.
Sure. Thank you. To the first question, it's a fair question. I think we said data centers are asset heavy, and we were not looking at it. We are not really departing from our core business model of remaining asset light. We see this as a logical extension. AI has made compute and data center capacity very scarce. Data center capacity and compute has become the strategic bottleneck in the entire value chain. Owning a foothold there will allow us to sell the full stack of what we believe is a higher margin services with the sovereign assurance to both enterprise clients and governments. In other words, we are making a targeted asset-heavy move precisely to grow an asset-light services engine in the AI era.
Of course, we continue to evaluate what are the right opportunities and how do we continue to expand our AI business, and this decision was as a part of that. The second question was?
Top five customers.
Yes. Sorry. The top five customers has declined because of one client where there is a year-on-year reduction because it's a large transaction that we did. After the first year, second year, there is a reduction, and that one client has significantly impacted the top five. But if you want to take as a broader sense, it was in line with what we expected, n othing unusual.
Our next question is from Poulomi from the Financial Express. Poulomi, please go ahead and ask a question. Poulomi, are you able to hear us? We'll take our next question. Our next question is from.
Hi, am I audible?
Hi. Yes, you are now. Please go ahead.
Yeah. Good evening. You've mentioned a strong deal pipeline, but I'm just trying to understand like what is the percentage of AI-led deflation that you're seeing on traditional revenue? Also, what is the reason for the dip in headcount, and slowed fresher hiring as well? Do you have a target in mind for your fresher intake for the year?
Yeah. So, AI deflation, we had called out. For our portfolio, we expect it to be 2%-3% is the AI deflation. But of course, in spite of AI deflation, our guidance is 1%-4%, which means we are growing much more to offset the AI deflation to still deliver a net growth. And the drop in headcount is a part of our overall business. We hire based on demand. We continue to evolve as an AI solutions company, which means the IP components, and the automation, and AI infusion in all our service delivery makes it easier for us to deliver the same work with slightly lesser number of people. If you see the last five quarters, our revenue per employee has gone up every quarter. So, that's really part of our overall business model.
Got it. Do you have a target in mind for the revenue per employee, since you've mentioned it?
We don't have a target, but we expect to grow it gradually.
Got it. Thank you.
Thank you, Poulomi. Our next question is from Rukmini from the Fortune. Rukmini, please go ahead and ask your question.
CVK, evening. Two questions on strategy and two on number. One, the Sarvam investment, if we look at the India share of business, about 3%, and you seem to be bullish on India business growing. Given the nature of businesses out of India, the margins, et cetera, the share of revenue from India business, is there some number that you can share on how big this is likely to grow because of Sarvam ? Two, on this data center business that you are talking about, CVK, if it is AI-optimized, should be about $1 billion for this 50 MW, right? Some sense of understanding on what kind of partnerships will there be, some equity partners stepping in investments that is going to be tied up? Perhaps once you answer, I will ask the second set of questions. Thank you.
India business, as we mentioned, the AI opportunity in India is at least $20 billion, and it is growing fast. While it is still a small part of our revenue, over the last year or so, we have been significantly enhancing our focus on India market. We plan to strengthen our India presence and also really lead with AI-native solutions in the Indian market, and that kind of builds the entire strategic rationale as well for our investments in a firm like Sarvam. Data center business, there are multiple ways of this business. One is a pure physical infrastructure or really a co-location kind of service. But the second one, you could be really providing compute and co-location. And the third is you can offer the models and the applications on top of it. We are going to play the full stack.
We have obviously data center. We are going to build compute GPU infrastructure. We will build embed models and lot of applications with both enterprises and government. And another very important use of these data centers will be to use for our managed services contracts with global clients and our fixed price or outcome-based contracts with clients. We can continue to consume our own capacity and deliver them much more cost effectively for our clients. In terms of the partnerships, in terms of the clients, in terms of very specific investment plan, we will share it as we progress, but o ur initial business plan is to invest about INR 3,500 crore. This will be a combination of debt and equity, and there could be other partners who will contribute to this as well, which we will announce in due course.
Got it. CVK, on numbers, the outsourcing cost, even on a year-on-year basis and sequentially has gone up. Some color on that, and two, also, s ure.
You want to take?
It has gone up, but you would have seen there's some reduction in the employees' cost, so it 's just compensating. We do use outsourcing cost to ramp up the execution part, so i t's normal in our business to have this both the cost, employee cost versus outsourcing cost playing together.
Yeah, the softness in the ER&D services business vis-a-vis the AI story, j ust want to understand what is happening in that bit if we are so bullish on the entire AI piece.
We had talked about discretionary spending being soft, and t hat directly plays into our ERS business, because we report it as a separate horizontal segment. So, you will see that a little more visible in our engineering services. It's very dependent. Once discretionary spending picks up, ERS will be one of the top growth drivers for us.
Thank you. Thank you, CVK. Our next question is from Rishabh Shah from Moneycontrol. Rishabh, please go ahead and ask your question.
Hi. First of all, congratulations, gentlemen, on beating estimates in a weak quarter. Two questions, CVK. TCS, one of our largest peers, recently announced, and said that AI revenue tends to be lumpy because many AI engagements last only for a quarter or two. Is that also what you're seeing at HCLTech? How are you working to make AI revenues more recurring or sticky over time? Second, CVK, could you please shed some light on how much of your deal today is outcome-based as it becomes more relevant in the age of AI and token maxing? Thanks.
Yeah. Our classification of advanced AI revenue has AI factory, physical AI, our IPs, which are on annual recurring revenue model, and semiconductor for inferencing, and things like that. It has got a good mix of annuity revenue and some project revenues. Annuity revenue would be like AI data centers. Then some of the recurring revenue that we get on software products, which are AI products like AI Force and other aspects, t hey will be recurring. Of course, there will be operating model transformation. There would be agentic AI implementation. They could be projects. It's a combination of both. Overall, this portfolio looks very healthy for us. It's growing. It has grown 60%+ year-on-year, and our overall aspiration is to grow 30% year-on-year on this business. The second question was? Sorry, could you repeat your second question?
If you could shed some light on how much of your deal today are outcome-based?
There are two dimensions. One is time and material and then output- based. And output- based, the way we look at it has both outcome and output. I think the last reported numbers were about 50/50. Within that, exactly what is outcome-based, we are not calling that out externally at this point.
Thank you, CVK.
Our next question is from Padmini from The New Indian Express. Padmini, please ask your question.
Hi. Good evening. Am I audible?
Yes, you are. Please go ahead.
Several announced deals involve AI factories, GPU infrastructure, right? How are margins on these engagements compared to traditional applications and infrastructure services? Revenue per employee reached record high and value headcount declined. How much of this improvement is driven by AI productivity versus workforce optimization? Third, you maintain your full- year revenue guidance despite record bookings, so w hat are the assumptions behind maintaining the current guidance rather than being more optimistic?
Yeah. Okay. AI factory margins, we don't call out such a granular level of details, but b eing completely new services with very niche capabilities, it's only logical to get much higher margins on this AI factory services than the traditional business. The second question is revenue per employee and the impact on AI. I think it's very difficult to call out the exact reasons, but a s you can see that our AI-led propositions, our AI Force, our entire automation and productivity and efficiency is really getting strengthened with the capabilities that we have. You should logically assume a large part of this is driven by productivity that we are getting with the use of AI in managed services contract. There could be some things which could be related to other aspects as well. The third element was assumptions behind the full- year guidance.
We continue to maintain the same guidance as what we shared in the beginning of the year. We have delivered a good quarter and our bookings are strong. Right now, our assumption is rest of the year will play out the way we expected, and that's why we have maintained the guidance.
Okay. Thank you.
Thank you. Our next question is from Anas from The Morning Context. Anas, please go ahead and ask your question. Anas, can you hear us?
Yes. Hi, am I audible?
Yes, you are. Please go ahead.
I wanted to ask, can you quantify the productivity commitments you're now making to clients when large contracts, for example, come up for renewal? Like on a typical renewal, what percentage reduction in overall project value are you seeing because of AI-driven productivity? Basically, the productivity gives back that you're seeing. What is the kind of percentage that you're seeing over there? Are you basically giving a full pass-through, or are you taking some of that into your margins?
It's very different for different kind of services. For software development, it will be like 35%-40% productivity. For infrastructure operations, it could be 5%, 10%, or 12% efficiency. In BPO operations, depending on the work, it could vary anywhere from 25%-40% improvement over a three-year or five-year period. So, it's very dependent on exact quality of work or nature of work.
Are you taking the full productivity? Are you passing on to the clients? Are you taking any margin out of it?
We are really not wanting to comment on that, but obviously, we have long-term relationship with clients, and there is a lot of dependency on the clients as well. So, it is very situational.
Okay. Thank you.
Thank you, CVK. Our last question is from Debangana from Moneycontrol. Debangana, please ask your question.
Hello. Am I audible?
Yes, you are.
Yes. Hi. I just have a couple of questions first to CVK. Your India business has been growing faster than other geographies. Wanted to understand what's driving this growth, and now that you are also venturing into sovereign AI and building capabilities around that, how do you see the India revenue contribution sort of grow in the coming quarters or maybe the next couple of years? My second question is for Ram. Want to get a sense of, since the company completed its restructuring, I think in FY 2026, so as of now, how does the workforce mix look like? Because I think TCS was talking about having equal number of AI agents and employees together working. How is the workforce mix at the moment changing at HCLTech?
Maybe let me comment on India business growth. Obviously, we are focusing on India, so this business is growing nicely. But this quarter, there is also a technicality which caused the India business revenues to be much higher, which maybe Shiv can explain.
Yes. As we report the geographies-based revenue, based on the customer location, we do the billing, we do the customer locations, and o ne of our global customers moved the consolidated billing and moved that billing to the GCC center there in India. That caused increase in India revenue this quarter, so nothing extraordinary to that extent. It's the business-as-usual revenue for this.
I think the workforce mix, it is probably too early to comment on how the agents and humans mix keep shifting. But it is reasonable to assume that as we leverage more of AI and we use more agents, the number of agents deployed in our work environment will increase, and t hat shows up in the productivity and hence the revenue per employee. As you see that trend move up north, you will know that the mix of humans and agents are shifting, and you will see more and more agents in the mix.
Just had one last question. Adding to what you had mentioned, CVK, about the West Asia war impact. I want to get a sense of how are the conversations shaping in West Asia? Are there delays in sales? Your peers have already mentioned about that, Accenture, TCS. How is HCLTech assessing that?
Our exposure to Middle East is very, very small. It's sub-0.5% is our Middle East exposure. We did call out some impact of this in the broader client portfolio in March, and probably some of that sentiments continue. I will leave it there.
With that, we will close the press conference. Thank you, everyone. Thank you for joining us.
Thank you.
Thank you, gentlemen.