Panel · 2:45 to 3:15 PM · Thursday, May 14, 2026
Our next panel will be led by, will be moderated by my colleague Ashwin, who's come from the other side of the world, all the way from India, and he's going to be leading this panel to discuss the future of IT services and business process outsourcing. And we're running a little over time, so let me just invite Ashwin and team to come on over. Hi. Thank you.
All right. Good afternoon, everyone. I'm Ashwin. I lead our technology services research here at HFS. So fellow panelists, our mission is to discuss the future of IT and business process services. And we have 30 minutes to figure out the answers. So it's a pretty straightforward mission for us. So let's do a quick round of introductions. So I'd like you to tell us about your background quickly. And then one word or one phrase that comes to your mind when we think of future of IT and business services, you're not allowed to use agentic or AI. Satish, let's start with you.
Hi, it's always great to be back. Satish Nair, I head up Global Client Relationships by Infosys. Change is a word. Change.
Hi, my name is Saju Joseph. I look after GBS BPO for Becton Dickinson. The word I would say is bright. Bright, okay.
I was at Citi for about 19 years, heading the GBS, the finance and risk share services and infosec operations, always based in New York. So I would say the current state, I wish I could say bright, but I would say conflicted.
Hello, I'm Jon Harding. As it says there, I was the global CIO of Conair Cuisinart, the consumer products company for quite a long time, but I prefer not to say long time. I prefer vintage or seasoned, you know. Or the experienced CIOs become advisors and consultants, right? Oh, and my word is different. Different. Different. Sorry.
Hi, my name is Ashish Chaturvedi. I lead BPO research at HFS. And the word that comes to mind is recomposition.
Okay. So it could be a different kind of change, recomposed, which can lead to a bright but yet conflicted future. Okay, we'll get there. So what we wanted to do is we have been thinking about how the future of IT and business process services will look like. And we obviously have been looking at various parts of operations and what the impact could be. So this is a very simplistic framework in the sense that what we are trying to say is, if you look at the bottom left, there are certain services which are very amenable to AI-led disruption. This is where the true Services-as-Software team will come into life in our view. You move to the right, there are certain set of services which would be hugely AI-enabled, but then humans will always be leading the thinking, leading the judgment. On the top left, there could be platformized set of services where it's the machines that are accelerating it, but there are humans to set policies. And then thankfully, there are certain set of services which we still believe will remain immune to AI, and we get to have our jobs, right? So I wanted to take some reflections. So as I said, Arindam, maybe we'll start with you. It's a simplistic framework, but your reactions, and do you see things spanning this way? Yeah, your thoughts.
Yeah, I mean, you know, while it's sort of simplistic in a two-by-two way, but there's a lot of complexity in each one of those boxes. But I would say we are seeing where we had expected the changes to happen first, which is at the bottom left, right? Repeatable process, it's relatively deterministic, it's rule-based, transaction processing, things like that that are happening, code generation at scale, right? So that's happening already. That was the very first kind of two use cases, as you can say, had been happening for the last year, year and a half. The blocker, so that's a good part, right? I think the top right is almost like, this is like meta task, you know, it's something that at the moment with the current set of technology or the capabilities we believe, only the humans are capable of doing it, but that's gonna change rapidly. The blocker is the bottom, the top left, right? Because I take, you know, the function that I ran for several years, this is all like pre-AI between 2014 to '17, '18, is the identity and access management for all of Citigroup, right? 8,000 applications for 220,000 employees, 100,000 contractors, 40,000 corporate clients. A lot of it, a lot of it can be done by AI. Yep. Absolutely, right? What humans needs to do is to set the policy and the exception. I'm literally reading your text because it makes a lot of sense. I'm not reading it just because it's up there. The humans' role, enterprise role, is to set the policy, set the exceptions, think about the security threat and do it that way, the risks and the frauds, potential and all of that, and let the AI run the place with obviously some human in the loop to make sure that it's meeting it. But we haven't gone there as fast as we should have. So that's why I think it's the block.
Fair enough. So, Saju, you said change, bright, right? So the future is bright, but there's obviously work to be done to get there. So BD is highly regulated, you know, there were a fair bit of acquisitions. So, and we've been talking about the foundational work that is required to really get there, right? So can you explain the thought process and the kind of initiatives you are undertaking BD ready for the future AI-enabled model?
Yeah, so if you look at a company like BD, 22 billion, maybe 45 billion products made a year across healthcare, connected devices, biopharma systems, extremely complex, 190 countries. So if you look at our problems, it's very hard to fit it into a four-blocker. Okay, so a company like BD, and companies like that, which are regulated and complex and have a heritage of many, many years. What you do is, from our perspective, you're building a framework that allows you to be dynamic. And that really has three parts. The first is, you really have to go after outcomes. And someone spoke about it earlier. It was really good. You know, people mistake, go after transactions or they go after capacity. It's really difficult to go after outcomes because they have different components of the box that you shared. And then the next process under that is really how do you change your processes? So we have a whole process where we use something called BD Excellence 2 to ensure that we quickly change processes and make them better because that makes it much more amenable for change. And lastly, something that nobody really talks about is tribal knowledge. We heard about it in the morning too. Within BD, there's a tremendous amount of tribal knowledge. And what we are doing is we are now ensuring that whatever all your communication is captured is sort of frameworked. So we don't really, we sort of define through templates and methodologies and processes how you capture your conversations, transactions, meetings. And I think that these three create a platform for change. And that's really, I think, what the stage we are in today.
Fair. Satish, you said change. And obviously there's a lot of talk about process change and reimagination as we think of the future. So what do you think the role of future business processes would look like, right, and what it means for Infosys BPM?
So it's great to see the data and operations as one small line item on the left hand side, you know, that bucket. The way we are approaching this whole thing is it's not a bunch of activities. These are functions that we actually run. These are areas across industry services and enterprise services that organizations are running. And these functions have a bunch of things that fall into each of those four categories which say, hey, listen, there is AI, there is going to be elimination. That's the way it's actually going to go through AI. But the ability to orchestrate this through a very fragmented enterprise ecosystem where the challenges occur at the boundary of a function, boundary of a process, and the boundary of the enterprise itself, right? So how do you bring together the change? How do you bring together the ability to get the function, stitch that function together, get the different skills in place for being able to manage this whole thing, manage control, compliance, right? Because you are trying to deliver to a very deterministic outcome in a very probabilistic world, right? So think about that as a way that you're actually landing it on the ground to deliver a certain compliance and outcome to the business. And that is delivered through these business functions. I think that's the way we are looking at that change.
Fair enough. Jon, would you like to add in terms of what this transition means? You obviously have an experience as being a CIO for many, many years. What does this transition mean? And how would you kind of add to what's been shared so far?
Sure. I mean, I think the transition we've all talked about over the last few days, there's a tremendous number of moving parts in this, and I think it's still a little bit about everybody being on the same page about we have to do this, we have to get into it, figuring out where to start. I think there's still an expectations gap in many organizations, both on the services side and on the consumer side of the services.
Sure. Will you come to that? What do you think needs to happen to close that gap?
Well, I think more common understanding, more talking, more transparency on both sides and partnership. I thought we had a great conversation earlier in the HFS contract, I forget the jargon, but the contract renegotiation discussion. I think that really boils down to are both sides coming to the table willing to change and willing to trust, as it were, and demonstrate. Sure.
I'm going to point that, but I'll wait for Ashish to make this call.
Okay. All right. So, Ashish, you said recomposition, right? And we always have this banter in terms of whether the death or decline of the services industry is exaggerated or not. Right. So, we'd like to take your view as an analyst. What do you think?
So, the reason I said recomposition, there are five elements to it. If you look at the building blocks of services, how you measure one unit of work, a single unit of work, first, second, the pricing construct, third, the talent pyramid, fourth, the buyers in the room, fifth, the competition at the door. I think all these five are getting rewritten at the same time. So what's essentially happening, it's the same set of atoms, but now they are forming a very different molecule. And that's why I said recomposition. On the point of whether services is passed is expiry, I think after every few years, somebody writes an obituary for the industry. And it's a very sensational but baseless statement to make. It has no merit to it. Having said that, I don't think it's hunky-dory out there. Just because we are past that hockey stick growth that the industry experienced for about 15 years, that's done. This is the first time in history that the headcount and the revenue are getting decoupled. So obviously these are ominous signs for the industry and the answer to that sign is in the quadrant that you had showed earlier. So if we look at the bottom left of that quadrant, which is essentially 25% of the market, repeatable stuff, AI executable stuff, transactional stuff, that is getting disrupted. But I would argue that the remaining 75% of the market has scope to become 2x the size with giving 3x the value. And I'll explain with an example. Let's take any example in BPO. Under FNA, we have accounts payable. Now how accounts payable is done generally is you get an invoice, you process that invoice, you pay that invoice, and finally the vendor takes a charge on it, a transaction charge on it. You repeat it million times or 10 million times, basis the magnitude of the deal and you get money for it. Essentially, that's the old model. Now, if you are going to remain in that old model, AI is going to eat you up. But the same vendor has now an option to manage end-to-end working capital for the client. It can model the impact of tariffs across 2,000 suppliers. It can also optimize the timing of making payments. It can detect fraud at scale. If I can think of four in 10 seconds, I think there are 20 more use cases available. Another thing is right now you have a vertical relationship with the CFO. If you are able to provide these services, now you have a horizontal relationship with the CFO, which is much more valuable, and which will gain you that. So according to me right now, what's happening is services industry is like wine. And it's the barrel aging phase which is going on. Now it has all the potential to deliver you vintage, but you might also get vinegar if it gets exposed.
Would you agree with Ashish? I think the vintage and vinegar is an interesting one. I think also sort of, you know, to paraphrase what you just said, I'll put it a little bit more, you know, in a funny way. It's like Mark Twain's famous quote, right? The rumor of my death is highly exaggerated. So I think the services industry is kind of going through that moment. But what I would also say that, for the enterprise, since I come from the enterprise side, if you lead this period of disruption with your providers and if you lead with contract as your instrument of choice, you're dead in the water. Absolutely dead in the water. We have heard that over and over again from a number of people yesterday, from Amit and Wipro, today from Mohammed, as emphatic as it could be, you've got to lead with the outcome. If we can't figure out the outcome so quickly because people haven't dealt with that, lead with the process. Contract is the last thing to lead with. Yes, Saju.
Yeah.
That will kill me for saying that, but go ahead.
As a consumer, I think a contract is really important. Legal visit, absolutely. But in the end, what's really changing, even a lot of discussions here is about are we, is services industry dying or living or whatever, but nobody will, we will always pay if we see value, whichever way it is sitting. And that's really all people are looking for today. So what you see now is, I see new people coming up with services, but services are still there because there's so much complexity in what we're trying to do and there's so much unknown. We can do it if we had the people, but we also have a constraint, budget, manpower, et cetera. So you need help and someone has to give you that help. It's not gonna come from a software right now. So the composition is changing, but whoever changes faster, better. And we see this, you know, we've been seeing this throughout industry. So I'm pretty confident that the industry will continue.
I would just add something I wanted to say before is, though I think from the enterprise side, there is an expectation that this is going to cost less. Just because that's the hype and we can't get the CEO, the board, past that easily. Yes, you can get more or at least get more for the same money. There has to be as part of this partnership that recognition. Otherwise, I think it's going to be very hard to sell on the enterprise side.
Yep.
Satish, your view? Yeah.
Yeah, I think the vinegar and the wine part is interesting. Either you cook them or you drink them. But you still need them. But I think more importantly, if I were to take a step back and look at the fact that there is business and business runs through a certain type of functions and how it gets done. I think for me the big change is the whole part of how it will get executed. The functions start disappearing. Will suppliers need to get paid? Yes, they need to get paid. Right? I mean, let's be clear. Nobody's going to work for free. Right? The question is how do you drive change in an ecosystem? You take an average organization that are 25,000 suppliers. Right? How do you pay them? Right? Driving the change and nature of skills that you need to bring in and how is it that you can augment with AI to drive that change becomes so absolutely important.
Yep. And I'll stick to you, Satish. So that's a fair point. And Saju mentioned that there are new service providers that are coming up. But I wanted to understand what are the new services that you see Infosys BPM kind of having the opportunity to now take to the market, which is becoming more relevant. So Ashish talked about the 2x opportunity and 3x value. So how do you see your portfolio and how are you thinking about what you'll position for clients?
I'm going to base this on what we are actually seeing on the ground in terms of where demand is actually coming from, right? First example is what we are seeing. For example, we are the largest mortgage service provider in Netherlands, right? We run whatever X number. But the nature of service is a full-fledged platform service, right? We run the stack, entire stack out there. Now, if you look at what's happened over the last 18 months we probably run now 75% of all mortgages and revenues. Right? Why? Because an earlier speaker said your ability to learn, execute, to drive, change, leverage, agentic AI, and run with it becomes absolutely critical. Take another industry service. Again, take something like insurance and what we are seeing across the insurance world. And interestingly, that's expanding into the enterprise world. If you look at the autonomous sourcing piece, now the problem statement that you're trying to solve is you have fragmented buying organizations in each of these areas. You're managing billions of dollars of spend, right? How do you really bring in an ability to consolidate leveraging what AI can do, which you could not even do that earlier. So your platforms like autonomous sourcing, where we are actually running the catalog, aggregating that spend, being able to take it out there in the market for it to get consumed by individual buyers at each of these. And we are seeing huge growth in those platform services across the board. Right, so that is the big direction shift that we are actually seeing, and that is based on actual data that we are seeing across the last year.
Yeah. So one question, I mean, there's a point of clarification, Satish. So I'm just thinking as you're speaking, I'm thinking about Ramp. Like Ramp had been an incredible success in the payable, you know, accounting, expense management space for a lot of startups and mid-sized corporates. How could they move so fast while some of our very large, because I know you're moving to the provider question, they didn't get there fast enough.
So there are two parts to it, right? One is just offering that as a product versus offering that as an end-to-end service across the globe. The ticket sizes that we're not structured to manage, for the lack of a better term, a mom and pop shop and go in there and ramp up the number of customers. That is not what we are structured for. We are structured for managing enterprise complexity at scale and be able to scale that and be able to deliver that across these organizations. So I think that's one difference that you see.
Yeah, so I'm building on that. I mean, there's been a lot, we obviously covered a lot about engagement models and all of that. So outcome-based pricing, everything is a complex topic, but what do you think the fundamental design principles need to be in terms of the engagement models going forward between the supply side and the market?
Yeah, so I think we kind of went into that territory already, right? So I think the engagement model, and I'm speaking here, as an enterprise, what I expect from the providers, because I've dealt with them for the last 20 years, and variety of the domain, is that you need to have that flexibility, because other people will eat your lunch, right, if that doesn't happen. And secondly, and this is true both for enterprise, like the GCC type setups, as well as the BPOs, and let's say in lower cost locations like India, Philippines, et cetera, because the same model, one is just captive, one is provider, is that we have to get out of this absolute addiction of the headcount. Yeah. I'm not talking about headcount against outcome. That's a different mathematical equation. I'm saying just culturally, my power is derived from the number of people that work for me. That is a fundamental obsession on which this industry has grown for the last 15 years, right? That is the, I wouldn't say the Kool-Aid, it's like the steroid that we have taken. And that's how people's career have grown. And people have gone from, you know, at a junior level to a very senior level at a fairly fast clip because they could add, just because of the labor arbitrage, you know, 500 more people, you know, every six months. And from something to become something. And you've figured it out. That is the only way to go. And now you're at a mid, like, senior, mid-level of your career. And that whole model had been put on its head. And you're looking like a deer lost in a headlight, right? So that, I would say, is a very important thing. And then just one last thing before I pass it on to you, is that it worked despite a lot of demand for innovation, maybe not AI-led innovation, but a lot of demand for modernization, innovation, automation, it still worked because of a completely different macroeconomic factor, which is currency depreciation. If I take 2012, just as a reference point when this whole GCC and this outsourcing boom kicked off at a huge scale, right, at a mass scale. And now Indian rupee has depreciated to US dollar by 49%. You can do nothing and it'll still be very profitable.
And I think we drank that cool. So, Saju, I'd want you to reflect on that. So as you think of BPO and operations, and we think of talent models, so what does the future of a GBS professional look like? And are there areas where you've stopped hiring or started hiring to kind of align to the shift?
See, I think it depends on where you are, where you're transitioning from. But if you look at services engine, you typically hired, as in using IT language, single stack skills, accounting, finance, HR, and that was what you hired. And it was a sort of tower-centric or vertical-centric model. That is changing to a horizontal model, where you're increasingly driving your value through horizontal skill sets. So that is reconciliation, that is some math skills, analytics, processing, and you're able to break it out. Now in theory that works, but there's a lot of outliers. So a lot of your hiring, and I would say it's not even hiring because it should be looked at in terms of your partners also. A lot of hiring is actually what you're actually building up with partners, but you are looking at building horizontal expertise rather than vertical expertise. The vertical expertise is getting much more specialized and they should be very consultative. So someone said earlier that you need cross-functional people or something like that, that is really true. And the second thing I see during this time of transformation is we tend to over-index with process talent. You actually need sales talent within the company because a lot of this is change management. And people who really drive, understand the transformation and are able to convince people to go along with it. We have very few salespeople within the company. Yep. And that's a problem. Yep, change agents, sales, yeah.
I'll get more hiring. That's one part of it, but you know, I just want to reflect on what Arindam mentioned. I think the industry ecosystem, in most cases, I think forced a certain behavior, right, to a certain point. We had a fantastic opportunity at the height of the hyper automation area where we could get away from a commercial contract model that fixed the equation of. Three times M, yeah. I mean, you could have easily done that. It's in our advantage to be able to do it, right? It did not, and I really hope that the focus gets more to the output and the outcome that you're actually delivering, right, rather than getting fixated again.
It counts, yeah.
So I can squeeze in two more questions, so different topics. Jon, I want to touch on the theme of trust. So there's trust between providers and enterprises, trust amongst employees when this transformation is happening, especially when you think of large organizational contexts. So what do you believe are some of the fundamental elements that should help maintain and build trust during such times of flux?
Yeah, I mean, this is the big question, right? How do we get to the change management that you talked about, Satish? And I think a lot of it is first belief that the technology isn't the thing. Don't focus on that. Try and focus on what are the outcomes we're trying to achieve. You know, start from the top. What is our business vision, whether we're on the services side or the buyer side? And work back from that to what are the key business capabilities we need to drive that? Where do we focus on? I think you can build trust by focusing on a few areas. You know, it's the old quick wins always build confidence and credibility. So, yeah, I mean, I think there is a long way to go on everybody on this.
It's a flywheel kind of an equation. Yeah. Yeah. And then finally, Ashish, I know I cover technology services, you cover business process, we have engineering services, data services. We saw the two-by-two in terms of how the disruption could potentially happen, right? Do you think there's a pecking order in your mind in terms of which areas will get disrupted more, will stay relevant?
So more than the pecking order for the area, I have a pecking order in mind in terms of who holds the key in terms of AI transformation for enterprises.
Okay.
And the research is clear. I have some bad news to share because there are providers here. But if you will peel into that bad news and you'll delve deeper, it's not that bad as well. And I'll come to that part. So we asked enterprises, who do you think is your anchor partner for AI transformation. And the first choice was hyperscalers, followed by technology companies like SAPs and Oracles, followed by AI specialist companies like Palantir's. IT services ranked four, BPO services ranked fifth, and then there were startups which were nipping at everybody's heels, which is number six. So that's the bad news, but when you talk to them, then they say, no, we appreciate what service providers bring to the table. That when there are headlines to be made, it's about the platform rather than someone who operationalized the platform. And you know, backstage we were discussing about airlines, so I'll give an example of an airline. So essentially hyperscalers are the ones that are developing airports. These technology companies are selling you airplanes. These AI native firms are saying we can provide you autopilot. But the problem is when you have turbulence at 35,000 feet, you require ATC. And that air traffic controller is the IT and BPO services providers. So that's why they are relevant. Another thing is, but what they can do to remain relevant, I think they need to take or capitalize on all the advantages they have right now. One thing is, who is above in the pecking order form partnerships with them. That's the natural one. Second is, you think about, there is a big journey from thinking about a model or implementing a model and gaining trust in that model. I think partners are positioned much better than a hyperscaler to do that. A third thing is SLS to outcome level agreements. That's what I call them. So that's another thing. And the most important thing is I think providers are best placed to rewrite the operating model for enterprises. A hyperscaler cannot tell you what decisions to federate, what decisions to centralize, what decisions to automate, what will remain with people. In fact, I have been in rooms where I know that the provider knows more about the client than the client themselves. So I think that's an unfair advantage that they have and they need to capitalize on that for making things happen.
I'll just add one thing, I know you're giving a point. I think the huge advantage that the BPO service providers have and probably have not been acknowledged for that for many, many years, and that's true for all people in enterprise operations, this is their moment to shine because they have the process knowledge. We heard that from everybody. If you don't have the deep process knowledge, this whole AI journey is going to fail. So bring that sort of in the context of the customer.
Context of what. So having everything coming from the world. Which is why you said bright. Yeah, coming from the world of consumer products, branding, this is a marketing problem. Everybody thinks of the hyperscalers because they're the ones everybody reads about and talks about. And the frontier models. You listen to, for example, we all listened to the gentleman from IBM earlier. I'd heard that story once before about their platform. What they've done is very impressive in terms of value. But when he said client zero, nobody identified with that. That's a marketing problem, not Chris and IBM marketing. I'm just saying, if you can better market those kind of stories that service providers tell, I think that would really help.
So we have figured this out in 30 minutes. Yes. Okay, thank you so much. I don't think we have time for questions. But thanks, anyways, thank you so much. Thank you.
Thank you Ashwin, Ashish, Saju, Satish, Jon, and Arindam.
If you don't have an account, Register here |
With the exception of our Horizons reports, most of our research is available for free on our website. Sign up for a free account and start realizing the power of insights now.
Our premium subscription gives enterprise clients access to our complete library of proprietary research, direct access to our industry analysts, and other benefits.
Contact us at [email protected] for more information on premium access.
If you are looking for help getting in touch with someone from HFS, please click the chat button to the bottom right of your screen to start a conversation with a member of our team.
