Operator
Ladies and gentlemen, good day, and welcome to HCL Tech's Q2 F527 Earnings Conference Call. As a reminder, all participant lines will be in the listen-only mode, and there will be an opportunity for you to ask questions after the presentation concludes. Should you need assistance during this conference, please signal an operator by pressing star, then zero on your touchstone phone. Please note that this conference is being recorded. I now hand the conference over to Mr. Nitin Mohata, Head Investor Relations. Thank you, and over to you, sir.
In the course of this call, certain statements that will be made are forward-looking, which involve a number of risks, uncertainties, assumptions, and other factors that could cause actual reserves to differ materially from those in such forward-looking statements. All forward-looking statements made herein are based upon information presently available to the management, and the company does not undertake to update any forward-looking statement that may be made in the course of this call. In this regard, please do review the safe harbor statements in the formal investor release documents and all the factors that can cause a difference. Over to you, CVK.
Thank you, Nitin. Good day, everyone. Thank you for joining us for HCL Tech's Q2FY27 earnings call. We delivered a solid quarter, and I would like to begin with the key headlines of the quarter. First, advanced AI services continues to scale rapidly and is becoming a key growth driver for HCL tech. Number two, new bookings were the second highest in our history and the strongest first half of any financial year. Number three, margins expanded significantly through disciplined execution and higher margin advanced AI revenue. Before I go through the numbers, please note that all growth rates that I mentioned are in constant currency except for bookings and ARR. Our services business, which is more than 90% of our business, delivered a strong performance. Services grew 3% sequentially and 4% year-on-year, with organic growth of 2.3% sequentially. IT and business services grew 2.7% sequentially and 4.4% year-on-year, while engineering and R&D services grew 4.1% sequentially and 2.1% year-on-year. Growth was broad-based across verticals and geographies, reflecting how closely our portfolios align with the clients, with what the clients are prioritizing today, be it AI, cloud, data, cybersecurity, or application modernization. Advanced AI revenue reached $219 million this quarter, up 28.4% sequentially and 78% year-on-year. And the annualized advanced AI revenue is now $876 million. This quarter's performance validates a strategy and advances our ambition to become the leading AI solutions company for the global enterprises. You can also see that we are able to offset AI-driven deflation by gaining market share in our core services and through our advanced AI offerings, which is helping us capture a share of the rapidly expanding AI spend across the technology ecosystem. Awesome. This is the growth template we aim to consistently replicate going forward. Overall, revenue grew 2.3% sequentially and 2.6% year-on-year. Software revenue declined 5.8% sequentially and 11.2% year-on-year, and this weighed on our overall growth. Headfield Software's ARR stood at $1.134 billion, up 7.6% year-on-year, which is the best measure of the health of this business, and it is growing. A little later, Shiv will share more details on the factors driving the software performance this quarter. Our EBIT margin was 18.2%, an improvement of 131 basis points sequentially and 71 basis points year-on-year, and it came ahead of our expectations. This expansion was driven by higher margins in advanced AI and gains from AI across our operations. We achieved this while continuing to invest in AI capabilities, IP, and talent. Let me now turn to bookings. New deal bookings reached $3.3 billion, up 27% year-on-year, and 35% sequentially. This is the second highest bookings quarter we have ever recorded and gives us our strongest personal bookings. Clients are working with us on more strategic programs and are choosing us to shape their technology agenda. AI was a central theme in every large deal we won. I will discuss them when I go through the pillars of our AI strategy. Our people count stood at 230,339 as of the end of last quarter, a net increase of 6,450 sequentially. Attrition was 12.8% on a LTM basis. Revenue per employee improved 2.4% year-on-year. A clear sign that our AI strategy is improving productivity across the business. I would also like to touch upon another topic. Over the last many years, we've ramped up our in-country recruiting efforts to build a significant local talent base. This is part of our talent strategy, and we will continue with this. In the last couple of years, the number of home applications we have filed is a very small number. Hence, we do not expect the U.S. Department of Labor decision to sustain the firm program to have any material impact on our talent plans or client engagements. Turning to our AI strategy, we continue to execute against the five pillars of our AI strategy, and this quarter, we saw them contribute as one integrated growth model. Pillar one is proactively transforming our services, where we are systematically embedding AI across our delivery model, and as AI adoption expands from targeted deployments to enterprise-wide transformation across engineering operations and modernization, we are seeing three clear benefits. First, AI is improving productivity, quality, and client outcomes. For example, for a leading private bank's complex core banking migration, we built an agentic AI solution that converts functional specifications into validated review-ready core for the target platform in about 10 minutes per routine with engineers approving every output. It averaged 87% accuracy for an estimated 9x productivity gain. Second, it is deepening client relationship and driving larger engagements. Guardian Life Insurance selected us to accelerate its AI-driven technology transformation and deliver seamless customer experience. In this multi-year partnership, Guardian will use HCL Tech's AI Force, our service transformation platform, to drive innovation across the enterprise. The third benefit we are seeing is AI-led transformation is now a key enabler of almost all our large wins. For example, a leading global industry technology company selected us to modernize its mobile application ecosystem. Using AI force and through a smart migration factory, we will deliver AI-led assessment, disposition, conversion, upgrade, and testing. This platform-as-a-service outcome-based model is expected to enable rapid and largely touchless modernization and support the transformation of 7,500-plus applications. Pillar 2 is building differentiated AI IP. As AI becomes universal, we believe differentiation will come not from models and compute, but from platforms, agents, and engineering IP that lets enterprises deploy AI securely, responsibly, and at scale. We have matured AI Force into a platform supporting software engineering, legacy and product modernization, data engineering, courts migration, business operations, and IT operations. We released a new version this quarter, and AI Force Agent Store now offers 900-plus agents, enabling reuse of proven capabilities and greater autonomy across enterprise workflows. Legacy and product modernization is emerging as an important growth picture. By combining AI force with our engineering expertise, we apply AI across code analysis, documentation, remediation, testing, migration, and support, lowering the cost, risk, and time these programs are traditionally required. This opens a substantial pool of client spending with adjacent opportunities in cloud data and operations. Our IP is also gaining recognition. A HCL DevOps loop from HCL Software, powered by AI Force, was named a visionary in the 2026 Gartner Magic Quadrant for DevSecOps platforms. We have built more than 20 industry AI solutions, moving us from horizontal capabilities to industry-specific processes and outcomes. And in physical AI, VisionX 3.0 introduced a fully agentic architecture with new releases for oil and gas and mobility industries. Finally, in our software business, we announced our intent to acquire RoboticIQ.AI, a Croatia-based enterprise RPA platform provider, to strengthen orchestration across AI agents and enterprise applications in HCL UMA. Our pillar three is expanding our AI services. This is an important growth engine and a strong area of momentum. The opportunity spans the full enterprise AI life cycle, defining strategy and governance, building compute foundations, our software IP, and deploying and operating agentic AI, physical AI, and industry solutions. Our agenting AI offerings have crossed $100 million per quarter this quarter. Clients increasingly want a partner that can integrate infrastructure, data, models, applications, and operations. Our AI factory proposition answers that need and enables HCL Tech to participate across the AI value chain, from compute infrastructure through the applications and business outcomes. Working closely with NVIDIA's AI factory specialists, our AI factory specialists were among the first and had built, validated, and benchmarked complete rack-scale GB200, NBL72 systems into it. Our AI factory proposition grew 64% QoQ and is a meaningful growth driver. An AI research company based in the U.S. selected HCL Tech as its strategic partner for a large AI factory program. The company is focused on radically improving AI data efficiency and making AI models mirror human cognition. HCL Tech will leverage its expertise in networking, cybersecurity, HPC infrastructure, and AI platform software to integrate and operationalize a production-ready AI factory. This then reflects an emerging category in which clients want a single partner to integrate, operationalize, and ultimately manage highly complex AI environments. A U.S.-based NeoCloud provider selected HCL Tech to build and scale its AI-native architecture and infrastructure capabilities that enable hyperscalers and enterprises to unlock the full potential of AI at scale. A Europe-based financial services provider selected HCL Tech to deploy an agentic solution to provide commercial relationship managers with real-time portfolio insights and proactive covenant monitoring, increasing capacity, reducing manual effort, strengthening risk management, and unlocking new cross-sell opportunities to support up to two times revenue growth. A U.S.-based healthcare provider selected HCL Tech to deploy ChatGPT Enterprise and establish a governed AI operating model that enables secure integration of AI capabilities across enterprise applications, delivering context-aware experiences that enhance employee productivity while maintaining enterprise-grade security, risk management, and compliance. A U.S.-based global power technology company selected HCL Tech to build an AI-powered digital thread that combines agentic AI and intelligent orchestration to connect engineering, manufacturing, supply chain, and service operations, enabling faster decision-making, accelerated time-to-market, and enterprise-wide AI transformation. There are several other deals in the investor release. Our physical AI continues to deliver encouraging revenue growth across robotics, IoT, and AI, building on our engineering and operational technology heritage. We are also continuing to find strong traction with our semiconductor propositions. A global semiconductor leader picked us to run a turnkey silicon development program from design through delivery for high-performance compute workloads aimed at the AI market. AI advisory is gaining traction. This enables us to help our clients at the start of the AI journey and then extend into helping clients transform their operating model and change management for an AI native world. For example, this is playing a key role in the large deal that we won with the Global Detailer and is currently under execution. In sovereign AI, demand is shaped by data residency, regulations, security, and national capability. In India, we announced our plans for our first AI data center in Bhubaneswar in partnership with Sarvam and government of Odisha. It marks a significant step in building the infrastructure layer of a full-stack AI strategy. Our fourth pillar is scaling strategic partnerships, combining leading technology innovation with our engineering depth, industry expertise, IP, and global delivery. This quarter, we achieved OpenAI Advanced Partner Status and joined its Cyber Defense Initiative, Today, we announced a global premier partnership with Anthropic to accelerate enterprise adoption of agentic AI. With a focus on safety, cybersecurity, and governance, we will establish a dedicated anthropic business unit and a center of excellence to accelerate deployments of cloud in enterprise environments. HCL Software will become an anthropic ISP partner, embedding cloud and cloud code into select enterprise products with frontier safety guardrails. We are utilizing all our ecosystem relationships to develop engineering capabilities and new offerings that create new sources of growth. The fifth pillar of our strategy is developing talent at scale, AI talent at scale. This pillar enables the other four. We believe leadership in AI will depend on combining AI skills with deep engineering, solutioning, and industry expertise. We continue to enable our employees to become AI builders, AI super users, and AI operators. We are scaling a differentiated forward-deployed engineer capability that combines deep AI engineering with customer discovery, solution architecture, and rapid prototyping, taking opportunities from use case to production. We are developing a centralized FDE academy with tracks aligned to OpenAI, Anthropics, AWS, Google, and NVIDIA. We are also investing in the next wave through an applied research program with MIT focused on safety and agentic systems and working with Microsoft to build a frontier transformation engineer program that further strengthens specialized talent aligned to strategic ecosystem platforms. Our focus now is to leverage these five pillars to create long-term growth through deeper adoption and investing accounts, new relationships through advanced AI, and participation in large transformation programs. Beyond these five pillars, we made several other moves this quarter that will support our future growth. We launched HCL Tech Pulse, a dedicated business unit for fast-growing enterprises with an annual revenue of $500 million to $5 billion, a segment we believe is undisturbed. It brings together our AI competencies to help these clients modernize and unlock value faster. We completed the acquisition of HPE's Telco Solutions Group, adding differentiated IP across OSS and 5G subscriber data management. We also expanded our India footprint. Alongside our AI data center in Bhubaneshwar, we announced a global technology center there to deliver AI solutions for global enterprises. We opened a new global technology center in Yip City, Gujarat, focused on financial services clients, and we launched our advanced semiconductor lab in Bengaluru with an investment of 185 crores, providing post-silicon engineering, testing, validation, and failure analysis aligned with respect to part to business. Each of these will contribute to a future growth. Now, coming to the pipeline and the demand environment and outlook, our pipeline remains strong and is quite balanced across service lines, verticals, and geographies. We saw healthy demand towards the end of the quarter. However, we know that the environment has been fluid and very dynamic, and the demand is influenced by various factors, so we remain cautious. As we see today, clients are prioritizing AI data and operating modern transformation whenever there is a clear case for efficiency, innovation, and business excellence. We have strong propositions to address these. As we look ahead, we are optimistic, driven by record first-off bookings, strong demand for our advanced AI propositions, improved margin performance, and our full-stack AI offerings that is being built incrementally. We will continue to invest the gains out of our operational efficiency to expand our AI advantage by investing in capabilities, IP, and AI talent. Our FY27 growth guidance is now 2.5 to 4%, and our EBIT margin guidance remains at 17.5 to 18.5%. Services revenue growth guidance is 3 to 4%. Our revenue guidance includes impact of our acquisitions. Shiv will provide you more details. Our confidence in this guidance rests on our strength of booking, a healthy pipeline, and sustained momentum in AI-led services. With that, I would now request Shiv to take us through the detailed numbers.
Thank you, Sri Rikis. Good morning, good afternoon, and good evening to all of you. Thank you for joining our Q2 Financial Area 27 Earnings Call. Let me walk you through our financial performance for the quarter. Starting with the revenue performance, please note that all growth numbers quoted are in constant currency unless noted otherwise. Total revenue for the quarter is $3,724 million, a growth of 2.3% quarter-on-quarter and 2.6% year-on-year. Services revenue for the quarter came in at $3,442 million, a growth of 3% quarter-on-quarter and 4% year-on-year. Our ITBS services grew by 2.7% quarter-on-quarter and 4.4% year-on-year. The ERS segment grew by 4.1% quarter-on-quarter and 2.1% year-on-year. HCL software revenue for the quarter is $294 million, representing decline of 5.8% quarter-on-quarter and 11.2% year-on-year. Let me provide some additional context on HCL software revenue performance this quarter. At the beginning of financial year 27, we had anticipated a software performance in the software business. I want to explain the reasons behind it. This year, and particularly in Q2, we had lesser multi-year licenses that were due for renewal compared to last year. The same is visible in the additional disclosure in the investor release this quarter. You would notice that the revenue recognized from multi-year subscription license has gone down from $81 million in Q2 FY26 to $39 million in Q2 FY27. As you would know, the revenue recognition for an on-premise term license varies based on the term of the licenses. Typically, upfront revenue recognition is around 80% for a one-year license and as much as 60% to 70% of the entire term for a two- to three-year license, with the balance recognized over the contract term. The reported revenue decline does not accurately reflect the underlying health of the software business. The same is evident in the healthy ARR growth that we have witnessed even in this quarter. Now, let me break out the inorganic revenue contribution in this quarter. HPE Telco solutions were consolidated for two months in Q2 and contributed approximately $21 million in revenue, helping services' sequential growth by 62 basis points. Daspersoft was consolidated for the full quarter, giving us approximately $15 million. Total inorganic contribution in the quarter stands at $35.6 million, which added sequential growth, revenue growth by 97 basis points. In terms of geographies, during the quarter, the USA grew 3.5% year-on-year, Europe declined by 0.3%, rest of the world grew 11.4%, and India grew 21.8%. Moving on to vertical performance, our Q2 growth was broad-based, with 5 out of 7 verticals resisting year-on-year growth. Growth was led by retail and CPG and financial services, which grew 14% and 7.1% year-on-year, respectively. In terms of clients, our continued focus on deep-think customer relationships has helped us expand the numbers of large client relationships. On a year-on-year basis, we added one client in the USD 100 million plus category, eight clients in the USD 50 million plus category, 10 clients in the USD 20 million plus category, and 33 clients in the USD 10 million plus category. Let me now share the details on the profitability. Our EBIT is at USD 377 million, representing 18.2% of revenue. Adjusting for restructuring expenses of 84 basis points, during the quarter, our Q2 EBIT margin was 19%, compared with 17.5% in the previous quarter. Net income for the quarter is $538 million, representing 14.4% of revenue. In terms of EBIT margin walk, there is a sequential improvement of 131 basis point in the company margin in Q2. The margin bridge explaining the theme is as follows. Project ascent, including AI-driven efficiencies, gave us 73 basis points. Superior margin of services offering under our advanced AI revenue gave us 32 basis points. Operating leverage at SG&A level due to growth acceleration gave us 50 basis points and Forex gave us 23 basis points. On the headwinds, there was 22 basis points impact on account of higher restructuring costs in this quarter compared to previous quarter and 25 basis points impact due to HP Telco solutions acquisition. Moving on to return on invested capital, ROIC. Our ROIC continues to remain strong, supported by our ongoing focus on profitability and efficient capital management. The last 12 months, ROIC is at 41.2% for the company, up 261 basis points year-on-year. Services ROIC is at 49.4%, up 411 basis points year-on-year. Exile Software ROIC is at 19.2% down to 64 basis points year-on-year. In terms of cash generation, over the last 12 months, free cash flow is at $1.78 billion, representing operating cash flow at, whereas operating cash flow is at $1.96 billion. Operating cash flow to net income conversion is 95%, while free cash flow to net income conversion is at 87% for the last 12 months. Our balance sheet remains strong, with growth cash at $2.66 billion and net cash at $2.56 billion as on 30th September 2026. Total DSO, including unbilled receivers, is at 84 days, representing a decrease of 2 days, quarter on quarter. And now for our shareholders, normalized diluted EPS for the last 12 months came in at Rs 70.22, representing growth of 5% quarter on quarter and 12.2% year on year. Including the one-time impact of new labor courts, diluted EPS was 67.56. The Board has declared an interim dividend of Rs. 12 per share to the quarter. The record date is October 16, 26, and the payment date shall be October 26, 26. This brings our last 12 months payout to Rs.60 per share, effectively distributing 88.6% of our net income. Now, an update on our guidance, FY27 guidance. For FY27, constant currency revenue growth is expected to be between 3% to 4% for the services business. And at the company level, the revenue growth is expected to be between 2.5% to 4%. The full-year revenue growth guidance now includes inorganic contribution of 62 basis points for services and 87 basis points for the company. This implies that midpoint of our organic growth guidance both at services and company level is broadly unchanged. EBIT margin is expected to be between 17.5 and 18.5 for FY27. That is all from my side for now, and I would now like to hand over this session to our moderator for a Q&A session. Thank you.
Operator
Thank you very much. We will now begin the question and answer session. Anyone who wishes to ask a question may press star and 1 on their touchstone telephone. If you wish to remove yourself from the question queue, you may press star and two. Participants are requested to please use handsets while asking a question. Ladies and gentlemen, we will now wait for a moment while the question queue assembles. Our first question comes from the line of Abhishek Pathak with Motila Lospal. Please go ahead.
Operator
You are audible, sir. You may proceed.
Yeah, hi, hi, hi, Sivike. Okay. Sirike, for us on the Anthropic Partnership, we also announced something very similar with OpenAI, you know, some time back. My question is, are we seeing some sort of structure being formed on AI implementation, taking a form similar to cloud migration, where ecosystem partners, you know, kind of had some benefit? And how is the value being shared between Frontier Labs and managed services right now? And let's say in the steady state, how do you expect this value to be shared as we implement and as we kind of integrate more and more of the AI Frontier Labs products or services into our offerings? That's one. And the second question is on our AI revenues. They seem to be accelerating quite handsomely. My question is, who are we competing with here? I mean, are these deals kind of, you know, more consulting led with existing clients, you know, Are we showcasing our own proactive initiatives or are they more computed a little bit on the market? And how do you see these deals going forward where their clients will have more and more vendors coming in to kind of gain their attention? And lastly, on the FCF bit, Shiv, there was some weakness over there on a TTM basis. I understand there's a couple of one-offs in the past couple of quarters. So, when can we expect to kind of get our cat cinemas and back to our three to four year average? Thank you so much.
Yeah, thank you, Abhishek. Let me start with the first question. You mentioned about AI partnerships with OpenAI, which we established almost a year ago. And the early days were a little bit trying to just understand and see how enterprises can leverage OpenAI, but now we are seeing a very good pipeline and deals for enterprise implementation of the enterprise that GPT and a number of other solutions that OpenAI has. I think our value here is largely in the EFDE motion, the services around it, and there is always a small piece of ongoing work because most of these platforms are still early days within an enterprise, so there is going to be a long tail of work that will come once we are inside an enterprise with one partner. I'm trying to scale. Anthropic is, while the partnership, we achieved the global premier status. We've been working with them for some time. And here again, it is largely the services model. But I think the biggest, the non-linearity is coming from the fact that some of our products, like AI Force, is getting integrated, and a number of software products are having native integrations into OpenAI, and now we propose to build it with Anthropic as well. There, I think we have a pricing which includes the token cost as well. So there, I think the equation will be a little bit not so much people-centric. It will be non-linear. So that's on the first question. And on the advanced AI revenue scaling well, there are certain components of this advanced AI services The two of the largest deals that we won, they're all single-sourced, primarily driven by our relationship with our partners. Now, and where we have come across some kind of competition, it is with some boutique providers. And there are some other components of advanced AI, like semiconductor, inferencing, chip design, where we see some of the service providers in the mix. The third category will be implementing an agentic AI solution, building an agentic AI solution. That is a place where we find a lot of service providers. And the fourth is all the IP platform that we build. We are very keen to, we have been continuously focusing on monetizing the IP while there is always a permission issue for a services vendor to go and monetize the IP. But I think we've crossed the hurdle of many of the large deals that we've signed. We have specific line item for a force platform. And then there is this operating model transformation of trying to transform the application landscape. While the application landscape infusing AI is part of our regular AI business, but the change management associated with it and the operating model of consulting, these two elements come under advanced AI, and we are building these capabilities as well, and a couple of large engagements have that. And I'll refer to Shiv for the EFCF question.
Yes. Thanks, CVK. Abhishek, the LTM calculation this quarter, you know, FCF2 and Iconversion has two one-time impacts. One is the BAPA tax payment, which is a bilateral advanced pricing agreement where we participate in government-to-government transfer pricing agreement. And linked to that, there was some tax payment. And the second is in the form of in-house, a shift from in-house to an external gratitude trust, which contributed to money being paid to the trust. Both these contributed to cash outflow in this LTM basis. So we should be able to see that this FCI to NI revert to maybe 100% plus stable as this one-off roll over from LTM calculation. I just want to add also that we have managed our cash conversion cycle very well, despite headwinds from expiry of Metcreds last year and lower non-cash amortization software business, which you would have seen in the P&L also.
Thank you so much. That was very clear. Thanks.
Operator
Thank you. Our next question comes from the line of Ravi Menon with Access Capital. Please go ahead.
Hi, thanks for the opportunity and congrats on strong growth and good margins. Susan, you talked about a 32-basis point contribution to margin improvement at Wednesday, I suppose if you look at it, it seems really high margins, so then it implies like high 30s and with margins for that segment on an ongoing basis. Should we think about it being that high, or is there also due to, say, some investments that you had maybe in people that, you know, that just you can think of as investments that now have got utilized and, you know, or maybe even investments in platforms where there was no corresponding revenue, but now we actually have the revenue and that's getting, the cost of getting spread over a larger revenue base. How should we think about that going forward?
So, yeah, there are two components. Yeah, go ahead, Shiv.
Go ahead. No, Shivika, please go ahead. Yeah.
Yeah, there are two components to this survey. One is the services component, and the second one is the IP and platform component. And IP and platform component, the margins will be very high, very similar to a software business. Yes. So the gross margins will be upwards of 70%. And for the services component, I think the margins will be at least 7% to 10% higher than traditional services. That is how we are thinking of this at this point. And that's how we see this play out as well. At least so far in all the deals, we have been able to generally do this at a better realization and much.
Great. Thank you. And, you know, just to clarify, this IP component is not classified as part of software product revenue, correct?
And one power up, you know, you have, you know, really good.
Just one clarification. Just one clarification. All our AI force and industry AI solutions, we still continue to classify them under services and then under advanced AI services as well. There are a couple of products in the software business which are completely AI native like HCL Uno, which is a universal orchestrated platform, and the VINIA data intelligence platform. These two are software product revenue, which are classified under Advanced AI.
Thanks so much. I'll just say on the guidance, the top end, I understand you wanted to maintain like a narrow band, and that's why the top end has been turned. But you're already at 4% year-on-year growth in your services business in PC terms, and you have a strong pipeline. You're saying that it's also good. Generally in Q3 anyway, seasonally, you don't have the weakness that your peers have. So why not really raise or keep the top end the way it was at 4 and out?
Yeah, I mean, you are right. We have a strong quarter on multiple fronts, and obviously, Kutrik will be a strong quarter due to the software step-up. But I mean, like the environment remains fluid. Like last year in February, we thought we were in a normal situation, and I met a lot of investors. And the day after that is when the Middle East conflict happened. Then a lot of things went in a different direction. So I think we are in a volatile world, and we don't want to increase our guidance and things like that. Things can change fast, both on the positive and negative side. We are seeing a good environment, and we hope it will continue. And the services, definitely, we feel confident to be slightly above the midpoint.
Thanks. One last question. For some of these AI deals where you talked about our clients are trying to adopt, maybe with a chance of pick or open AI, I would assume that there is no real TCEA attached to some of those, because these are more on an explorative basis and maybe on a time and material basis. Would that be right?
I mean, it's a mix of both. I think the whole change management-related OMT, which is Operating Model Transformation, that we have signed a couple of engagements on a fixed-price basis as well. But agency implementation on Google and other things largely comes as time and material.
Perfect. Thanks so much, Mr. Plot.
Operator
Thank you. Our next question comes from the line of Rishi Junjanwala with IIFL Capital. Please go ahead.
Yes, thanks for the opportunity. A couple of questions. One, our top client buckets across top 5, 10, and 20 still don't seem to be growing. I know you had called out in the past that there were three clients where there were, you know, rammed downs. But just trying to understand, are those ramp-downs behind us, and when can we expect these buckets to start growing?
Rishi, I think whatever challenge we kind of highlighted in the three of our top 20 clients, we had already factored that in our head, and two telco customers where discretionary spending was under pressure, I would think, I mean, whatever we had in mind has already played out, whatever we had assumed that has played out, but it's still, I mean, you cannot be sure because there is a lot of discretionary spend in the two Delco clients. But what you should see is, despite these headwinds, we are very pleased with the upselling to our existing customer base, as is reflected in the addition of 100 million, 50 million, 20, 10, client buckets on a year-on-year basis. So I think whatever assumptions we had, that has played out. I hope that answers.
Yeah, thank you so much. The second question is on margins. I mean, I just wanted to understand, I mean, how do we, you know, read the margin movements across services and the software businesses? I mean, in this quarter, we have seen, and if we look at even on a YOY basis, a fairly sharp expansion in services business margins. Software, on the other hand, has declined sequentially and still there is an expansion. So, just wanted to understand how do we, you know, read the divergent performance in margins versus how revenues are going?
I think in the Q1, we had one of restructuring cost impact that impacted the margins in Q1. And in the absence of that restriction cost as well as, you know, the benefits we got because of, you know, optimization in the employee-related expenses along with lower amortization has helped us to, you know, in the sequential margin expansion in this quarter. So the plan for software is we continue to plan and rationale the cost to align it with the growth outlook we have in software. So that's how you would see the margin expansion in software. On the surface side, I think it's just, as I said, the project ascent as well as, when I explained the margin walk, helps us as well as the uptick in the advanced area revenue that has been helping us to improve.
Operator
Thank you. Our next question comes from the line of Surindra Goyal with Citi. Please go ahead.
Yeah, thanks. So, just a clarification on the software business first. So, the organic decline seems to be like 15-16% year-over-year. So, what is the outlook here? The ARR would anyway also have benefited from the Jasper Soft acquisition?
Yeah. So, first is excluding the impact of Jasper Soft. Our ARR has grown 2.6% YOY, despite a 15% year-on-year decline in revenue. I mean, all of this is in constant currency. So, I think ARR is the right metric. I would request KK to give you a little bit. KK is the president of our software business. I would request KK to provide a little bit of outlook on the software segment.
I said earlier that the revenue decline does not actively reflect the underlying factors of the software business, and as CVK just alluded on the ARR number, it is a metric on which they are trying to operate the amount of multi-year renewals that were available compared to the previous quarter, but I think a lot of these renewals will come back in FY28 and FY29 as part of our renewal. On an overall mix of the software portfolio, we don't see much impact of AI in this, except we have a small footprint of application software or application software products, which is in there. Over the last few years, we've been pivoting on two big areas in our software business. One is what we call as intelligent operation. This is around products around IT operations, cybersecurity, and business process operations. The second is revamping our modern data stack, and modern data AI stack is around the AT&T data intelligence platform. And a lot of our acquisitions and inorganic, which we have done, capabilities that are built organically, including the likes of Xenia, Wobi, and the recent robotic, which CBJ mentioned in the start of the call, these are all feeding into the direction where we want to grow the business.
Thanks for that. Just a clarification, like last six quarters, I am looking at the data, the ARR has been growing. So I am just trying to understand how long can this disconnect between ARR and the reported growth continue. Yeah, so I think if you look at the revenue profile, there is two components which can be fluctuating. One is the perpetual license revenue. The second one is the impact of upfront revenue recognition of a multi-year license. I think, as Shil said, almost 60% of a three-year license will get recognized upfront. So that can cause some kind of cyclicality in this. Other than this, I mean, if you really take out the multi-year piece, then there should be a proper correlation between ARR and revenue.
Understood. And just one more clarification. Is there any seasonality in the HPE CTG business because the run rate of revenues seems to be lower than most expectations?
And it's for the two months, CTG you're talking about, we have not called out CTG revenue. The only thing we called out is the acquisition, which is Tempo Solutions Group of HPE, which is a recent acquisition. It's called Arobat ES. That had two months impact. So you will have to, and I think Shiv called out the impact of that. If you kind of extrapolate it to three months, it should largely give you the right number. There could be some small seasonality, but nothing major. Sure. Thank you.
Operator
Thank you. Our next question comes from the line of Gaurav Rateria with Morgan Stanley. Please go ahead.
Hi. Thanks for taking my question. My first question is on the outlook. It looks like there is a sharp deceleration in the revenue growth in services business on a sequential basis. In the second half compared to Q, if you look at 2Q, organic growth has been around 2.4%. So, is there any higher than usual scalability we are building? Is there any slower conversion of deals into revenue that we are building in? Or is it just conservatism that we have kept in our day?
It's called out when we announced the mega deal. It's not going to contribute to any revenue in Q3 or Q4. It is going to come in Q1 of next year. So, that is one outlier. So, other than that, we don't see anything to be very different. And generally, we've been conscious not to kind of create any expectations of higher growth. I mean, the environment has some uncertainties, and it's been a volatile last few quarters, so we have been cautious there. And services most likely will be better than the midpoint.
Got it. My second question is on the partnerships. I think that the ecosystem partners that we have is a very significant contributor to our revenues. Over the years, that number may have increased. Is there any sense of what that percentage of revenue looks like? And keeping that in context. Sorry, we couldn't hear you.
We couldn't hear you. Can you repeat your question?
Sorry, am I audible now? Yeah, yeah. Yeah, I'll just repeat my question. The ecosystem partnerships have contributed significant amount of revenues over the years, right? It could be a very large number. It would be great if you could highlight what percentage of our revenues come from our partnerships right now. And keeping that in the context, how do you think the partnerships with these frontier model companies like OpenAI or Anthropic would translate into material revenue from, let's say, next 12 to 24 month point of view?
Yeah, we are not very precisely tracking all the revenue that comes through the partner channel because, I mean, a lot of things that we work in our client environment, we also take it to the partners to scale different solutions and things like that. I know there is industry players calling out how much is coming from ecosystem partners, but I don't think we have a renewable mechanism to track it. A lot of it could be what we call as influenced revenue, rather than really a revenue coming from like if we are migrating somebody from, let's say, on-premise to, let's say, Azure, that will be treated as an influence. We have a new for Azure, and that is one measure that they use. But we obviously are focused on services, the annual revenue that comes out of it. We are not directly tracking that. And there is definitely one channel where we are operating as a professional services partner for almost all the cloud hyperscalers, and now for the frontier companies and the semiconductor companies. I think that's still a small number in the overall scheme of things, but growing very well.
Got it. And last question is your comment, Sivike, you made on margins that operating efficiency, whatever you get, you will be pulling back in the business. So what's the broad philosophy on the margins, right? Because a couple of years back, we had an aspiration to get to 1920, and then we were in the investment phase. So we brought down our overall margin number to 17.5 to 18.5 in that range. And now, obviously, you know, we had some restructuring charges this year, which I am presuming may not be a sustained charge for the next year. So keeping the puts and takes, you know, variety of tailwinds that you talked about on the advanced AI being higher on the margin side and multiple investment that you have kind of outplayed for yourself. What's the range that one should think about from a medium-term perspective?
I think our year aspiration was 18 to 19, and last year we had, I mean, this year we specifically called out the restructuring charges, and then we had reduced the guidance to 17 and a half to 18 and a half. But 18 to 19 is our desired gun, but I wouldn't want to kind of comment on it now because the industry is changing significantly, and we would always want to invest to gain more capability and differentiation rather than optimize margins in this space of the industry.
All right. Thank you so much.
Operator
All the very best. Thank you. Ladies and gentlemen, we will take that as our last question for today. I would now like to hand the conference over to Mr. C. Vijay Kumar, CEO and MD, for closing comments. Over to you, sir.
Thank you, everyone, for joining our earnings call. We've delivered a strong quarter on multiple fronts, and we have a strong execution plan, obviously our strongest quarter in any year, and we're all geared up to execute well and deliver a good outcome. And our guidance is also well informed by all the things that are happening, both within the business and in the external environment. And as I close, I wish all of you a very happy festive season ahead. And thank you, and good evening.
Operator
Thank you. On behalf of HCL Tech, that concludes this conference. Thank you all for joining us. You may now disconnect your lines.
