Transcript
Speaker: Competition can come in terms of how innovate innovation happens in their product set and in their models, right? So, you know if I went in the frontier model, my endeavor is to say, okay, how do I improve the model so that I can start eliminating the human effort?
Speaker: So you continue to push the boundary of innovation to you know to get better productivity and leverage the better ah intelligence capabilities that exist within the model. But when it comes to applying the context, they still need a partner like us.
Speaker: You're listening to From the Horse's Mouth, intrepid conversations with Phil First.
Speaker: Hello and welcome to the latest edition of the podcast from the horse's mouth. I'm your host Phil first and joining me today is another leader in the services industry.
Speaker: A gentleman called Venu Lambu. Hi, Venu. Hi Phil. Could you maybe yore the tell our listeners a little bit about your background and how you got to what you're doing today?
Speaker: you know It's been a while that nobody asked me this question about the background, so let me start from where my career started. Before that, Phil, nice to be here. Thank you for inviting me. You know, I started my career way back with a company called Digital Equipment Corporation.
Speaker: I'm not sure how many of... them remember that company. It was founded by a gentleman called Ken Olson, who ironically believed that the PCs are meant for home, not for businesses. And as it happens in the tech world, a PC company called Compaq acquired them in late 90s. So, you know, I've been through that, mean, from computing era to client server computing era.
Speaker: Then I was leading the cloud and technology business in Cognizant for a very long time. So, you know, navigated the The cloud computing era, and of course I was in Minetree, was on the board of Minetree before the merger.
Speaker: And then happy to be back, and almost two years back, I joined back LTM, which was known as LTM entry at that time. So let's talk a bit about what's happening in our industry right now. you know We talk a lot about ai revenue overtaking traditional IT services.
Speaker: And you're right at the core of this. And what truly has to happen for AI? these AI native revenues to become a material part of the business and and what happens to a traditional services revenue as this transition accelerates. Maybe you can share a little bit about what you're seeing here.
Speaker: you know, firstly, I look at AI in two dimensions. One is AI infusion into the core and traditional part of our the business. know, as I keep saying this, if you want, if somebody asked me, know, in simple terms, where have we been generating revenue all these years? It has been either we were building systems for clients or modernizing their systems, or we would be running the systems for them, application, infrastructure, security, and so on.
Speaker: So, you know, there is that dimension where the AI can influence all the work that we have been doing over the years. You know, that will make ah most of these services to be reimagined at a different price points. And we are seeing that already.
Speaker: Right. So the price per unit is anyway going down for developing products, for running systems, for modernizing systems and so on. So that's one area of it. And I do believe that there is still a lot of opportunities in that area.
Speaker: because the clients are still in a modernization journey in so specific sectors. So that's one area of opportunity, how you can get ai at the core of it, infuse, enable it, and and reimagine the core services.
Speaker: Then you have the the second dimension of ai services, which is about how do you you know, sort of reimagine the business process. How do you impact experience, supply chain, you know, the client acquisition processes or the horizontal process of HR finance and so on, the enterprise, so that the role that you can play in actually transforming the client enterprises into an agentic or a self-learning enterprises, right? Self-learning autonomous enterprises. So that's a huge opportunity that's out there.
Speaker: That's where the, you know, your concept of services as a product comes in because you can productize lot of the business processes and, you know, start creating a self-learning loop ah kind of solutions for the enterprises. So that's the second part of it.
Speaker: The way we look at that piece of work is broadly around three types of ai work, bill right? One is about business AI, which is what I spoke about, reimagining the business process, agentifying them, creating a more, you know, SLMs, you know, for specific functions and processes and so on.
Speaker: The second is a marketing AI, right? I mean, we are seeing a significant work unfolding in the entire marketing stack, right? From creative content to the entire technology stack of marketing, including marketing operations and how do you look at client loyalty and so on.
Speaker: So that that's the second piece of work. And the third is industrial AI. All the work we can do in you know, reimagining the manufacturing system, the design system, the supply chain system, all that is under industrial AI. So these are the three you know, services that I see as the next phase of growth, broadly classified as AI native services.
Speaker: And of course, you have AI led software engineering, which underneath everything that you do in terms of product development for customers. Yeah, I mean, we can look at this at the technical level where clients will reach out and want some assistance in areas where they need access to skills and capabilities. But in terms of the broader AI agenda, you know, we're seeing a lot of enterprises really struggle to come up with a unified strategy, really figure out how to execute and really work across the C-suite because this is so much of a business transformation as a
Speaker: as a tech transformation and we see firms like Palantir in the market who only target CFOs and COOs and CEOs, right? So how does that impact the role of a traditional IT t services firm who might not be used to working with such disruptive business transformation? Is it a very different conversation that you're being pulled into? Is this a rapid learning curve that you're being forced on?
Speaker: Absolutely. It is definitely a rapid learning curve for our sales force as well as for the delivery organization. Because as you rightly called out, you have more than one persona of buyers in enterprises because What AI has done is it has democratized the access to that level of intelligence.
Speaker: You know, it's cuts across all the processes, whether it's the functional processes or the sector-wise industry-specific processes. And there are a different processes for each of them. You need to have a conversation with CMO to reimagine the marketing you know side of the business. You know, you should be able to have a conversation with CFO to re-look at the entire business.
Speaker: you know, shared services, of the financial services finance part of the office, or you should have a conversation on manufacturing it and so on. Absolutely. So that's a big learning curve. And if I look at what we have done is I've logged on lot of salespeople in in one place, you know, every quarter and say, okay, now let's just create a lot of possibilities that we can deliver for the customers. So through those interactive workshops, we are actually simulating various scenarios and, you know, and then trying to take those conversations to customers and co-create those solutions in our AI studios and so on. But absolutely, from an industry standpoint, shift in how we sell and how we're going to deliver. And most importantly, how are we going to be measured in terms of the charges that clients want to pay us or what defines the success as well in terms of outcome and and output.
Speaker: There's a lot of talk around and decoupling people-based services and moving to this outcome-based subscription and consumption models. But can can this industry generally decouple growth from headcount in your opinion?
Speaker: It's not going to be easy because I think, you know, okay, let me put it this way. Is that the future? The answer is absolute yes. Right. I mean, we do believe that the business model has to decouple effort from the output and outcome that, you know, that that we can deliver to the customer.
Speaker: But the change to do that is not just one dimension in the sense it's not just the responsibility of a service provider. I think the whole ecosystem also has to rally behind it. You know, when I talk to the clients,
Speaker: You know, probably the CIO wants to do that, but the sourcing organization still, you know, looking at, okay, you know, how much of the risk gain is there?
Speaker: Are we getting a better deal of it? Or can we construct the contract differently? Or is there a contract maturity that we can arrive at? So it's a learning curve even for the sourcing organization, for the CX organizations, you know to sort of pivot everything to an outcome base. So it's going to be a gradual process, I would say.
Speaker: In some places, it will move fast. In some cases, it'll probably move in three or four steps. And by thereby, you start having an engagement switch, you link it with outcome-based construct, and and thereby, you start making an effort to de-link the effort ah with the work that you deliver to the customers.
Speaker: you know Having said that, if I look at over the last six or seven quarters trend, Phil, I'm already seeing in our revenue, right? So you know every quarter we are growing. you know In some quarters, we grow faster. In some quarters, we grow modestly.
Speaker: But in all scenarios of growth, I'm actually not increasing the headcount, right? So the headcount is either flat. In some quarters, it has actually declined, except the freshers that, you know, who we continue to hire to build and ah good foundational skills for the future.
Speaker: Otherwise, we we are seeing early indicators of efforts being delinked to our growth already. Is it happening at scale? Probably it's not dead there because I said it requires a whole ecosystem to get aligned and some standards to appear in terms of how we engage and how we measure.
Speaker: As we start to think about this shift from maybe cost obsession to performance, do you see the the translation difficulties happening between the C-suite and the sourcing organization. Because we see so many deals where the sourcing organization gets involved, they're reaching out saying, we'd like to see 30, 40% reduced because of AI, right?
Speaker: And there's little beyond that. Are you getting pulled into a lot of client conversations where they're just demanding cost takeout because they think you can just drive a lot of agentic into the services? Or do you are you trying to avoid some clients who are just looking at it in that way?
Speaker: ah You know, in terms of conversation and expectation, you're right. um I think to some extent it is coupled with some of the old stated narrative about the technology innovation that's available versus actual deployment structure that we need to adopt, including change management.
Speaker: and let's Let's take an example of software engineering, right? You know, the software engineering aspect, we see a high degree of productivity that's visible when we use the AI tools and technologies. So that has reached a good level of maturity, I would say.
Speaker: But to actually get the net productivity gain, it's not just about how well the service provider will do their part of the work. You also need to do the change management within the client organization, including the reskilling of the client you know talent that exists within the client ecosystem, because both the teams need to work together. Either you're doing a product development or you're working on in a new platform, modernization and so on.
Speaker: So there is that talent transformation aspect also that is involved. So my my message here is that whenever I get pulled into this conversation is that, look, Commercial is one side, right? So if you want us to deliver to a particular price point, that's a more of a strategic discussion. That's how we do it.
Speaker: Let's have a conversation in terms of how do we execute this, right? Otherwise, we'll be oversimplifying by just looking at 40, 50% productivity and not double clicking.
Speaker: How do we get there? So for me at the moment, the worry is not much about the price. The concern more is about as an industry, are we doing a good job of double clicking the execution steps?
Speaker: Are we good doing a good job of helping clients to deal with the change management? It's not a one step, you know, a switch or it's a one click bait where we just give a discount and, you know, we have achieved the AI journey. It's, you know, it's ah it's a journey that we have to travel together. So I would love to have a lot of conversation on the journey.
Speaker: you know, the commercials and all will fall in anyways as a part of it. But we shouldn't start with that as a starting point, is my view. Let's talk a bit then about the frontier AI firms like Anthropic, AI, Google. They're moving deeper into the enterprise, but do they ultimately become partners, technology suppliers or competitors to traditional services firms?
Speaker: At the moment, I think we are seeing a good partnership approach move from most of the frontier companies. We ended up announcing partnership with but with most of the frontier companies. So we do we do see that partnership in enabling our talent and skilling the talent and so on.
Speaker: You know, the competition can come in terms of how innovate innovation happens in their product set and in their models, right? So if I went in the frontier model, my endeavor is to say, okay, how do I improve the model so that I can start eliminating the human effort?
Speaker: So you continue to push the boundary of innovation to, you know, to get better productivity and leverage the better intelligence capabilities that exist within the model. But when it comes to applying the context, they still need a partner like us.
Speaker: but As the adoption intensity increases at the client organization, I think we will start taking a lead in that conversation. In the initial phase, it was the model companies or the you know platform companies which took the lead in conversation with the customers to explain them the you know the level of intelligence that can come in, the capabilities in the models and so on.
Speaker: But as the adoption velocity increases, we have a good seat on the table there to lead that along with our partner, of course, you know for our clients. Now, when you're starting to hear though that they're calling for better regulation, they're calling for more oversight,
Speaker: Is this because they're concerned of they're trying to protect their own position in the market? Do you think or does genuine, genuine fear that bad things are going to happen to us?
Speaker: You know, sort of take an approach that any any so technology that deals with data, business workflow processes, and anything that takes the responsibility of delivering outcomes to the businesses,
Speaker: needs to be regulated because there has to be some accountability for it to work in a certain way and to deliver it a more compliant way because all these businesses are controlled with all the regulations. So as most of the businesses are sort of regulated in some way or we have to adhere to the compliances, we have to make sure that any technology that is used in the client environment also has that framework which governs.
Speaker: Now, whether you regulate model companies or you regulate during the adoption is a matter of debate, because if you regulate the model companies, then you're essentially regulating the innovation. But if you're regulating how it gets implemented, then it's something we have done in the past when the Data Production Act came in, when the cloud came in, you know, the regulation came in due for the adoption cycle.
Speaker: I think that is pretty much needed, I would say. On the innovation, you can deb debate both ways. But I think on the adoption, definitely there has to be as sort of ah some kind of a framework on which we govern and there's an accountability for it. right So we don't want when we know that this technology has a habit or or has a potential of either going a bit rogue or going on a self-learning mode and trying to do things which you don't want us you know the enterprises to do it, then you have to regulate it as well. Yeah, I mean, it feel it feels to me like there's an opportunity now to help clients create guardrails between what moves onto a shared frontier platform
Speaker: what moves on to an open weight model versus what is true business context that needs to be sovereign to them right whereas if you're just servicing the movement to work out to these frontier firms eventually you'll be commoditized right so to what extent do you feel enterprise context becomes a new premium and then how does a company like ltm exploit that you know the context will always be a pre premium Phil, in my view, right? It has been a premium before. You know, if you look at even in the pre-AI era, if I call that, we got the deals and engagements with customers because either we demonstrate the the context of the client's business, whether it's a process understanding, domain and so on.
Speaker: Or we grew the account because we continue to appreciate the context, you know, that is needed to deliver a particular service. So it was always, it always had the premium. And I would say it's going to have, you know, if not more, the same amount of premium when you're implementing AI technologies to the client environment.
Speaker: Now, what are we going to do from an LTM perspective to make sure that the context capabilities are brought in is if I look at across the entire AI stack, we try to bring the context in every layer of it.
Speaker: If I'm talking about the AI assurance and AI security, you know, we need to bring the context of the sector at which the client operates in. the regulation certain processes that needs to be followed, and the vulnerability that needs to be checked for any of the agentic intervention that happens in any of the, that floats across the system of record, a system of engagements.
Speaker: So that's one example on how do we bring more context you know to that layer. Same thing when we go into an opportunity of creating an SLMs for the clients. right I mean, there are the the number of discussions we're having with clients is actually increasing about how we can create SLMs because every task doesn't need a very big frontier model. So you can actually create, you can turn out SLMs in six to ten, eight weeks, you know, which could be a self-learning autonomous for that particular function or the particular processes.
Speaker: So again, to do that, you need to bring the context, either a functional context or the domain context into that. So yeah, it is it is hugely a premium thing and I would expect it plays a role across all the stacks.
Speaker: the risk the clients are taking. In a weird way, they're taking less risk partnering than trying to do too much of this in-house because I've seen clients already have the little fiefdoms of agentic knowledge, leave the company, and it's very hard to replace. I mean, its it isn't like taking up, up, up, replacing them with more of them, right? it's ah It's a whole different talent model. So how do agents change the whole hiring team structures within it within a large company like yourselves? Are you already making shifts to flatten the organization around like experienced doers and and freshers? How are you adapting to this model?
Speaker: When we look at the talent strategy in the AI era, I took a very conscious decision saying that, look, let's not create two worlds that, okay, the existing team will do this. Let's go and build a new team to do the AI.
Speaker: In my view, we'll end up creating, you know, two different worlds, you know, and especially in a large enterprise like ours, we can't have those kinds of islands working in. So the first fundamental thing was that how could we make sure that the developers that we have out of 90,000 employees, let's say 40, 50,000 are developers.
Speaker: How do we make sure that they actually get access to the best of the AI tools and technologies and you know take them on a journey of enabling them on the AI skillset? So that process, I think we've pretty much done a lot on that area.
Speaker: We'll continue to do, but I think now it has reached a stage where we can deliver some serious work for our customers. So that's one part of the talent transformation that take the existing team along with you reskill them, give them enough access to the tools and technologies. That's why we invested a huge amount of money, both on co-pilot, GitHub, Anthropic, Cloud and so on, primarily with an objective of giving access to our developers and so that they can start delivering work for them. Now comes the second part of it, which is in terms of, okay, now how do we build the lateral hiring strategy
Speaker: you know, for the current and the future demand. And also, how do we get the freshers coming in? I'm really excited about the freshers cohort, right? we're We're doing high train and deploy model as well in the freshers cohort, apart from just hiring them fresh from college. We partner with them at ah you know at the final stages of their graduation and train them and get them in.
Speaker: I see a big change in their mindset. They are pretty much an AI native mindset cohort that you get. So I'm really excited about that fresh talent end that's coming in. Then you have the experienced talent that we need in a very specific areas, right? Whether it is the FDs kind of cadre that we are talking about or the exponential engineering cadre that we talk about or the harnessing engineering capabilities or, you know, the consultants who can bring the good tokenomics capability to do it. So we are doing a very targeted hiring and competency building in those areas, which doesn't get addressed by just reskilling.
Speaker: So hence the the combination of reskilling, freshers hiring, and very focused hiring on the gaps that we need to build organically, you know, is helping us to transform. And eventually I would say that,
Speaker: The pyramid structures will disappear. Well, to be honest, I really need less managers now. than a span of One thing that I advocate to my teams is that don't ask me the span of team.
Speaker: Tell me what's the span of work that you manage. Right? So this the suddenly the measurement para of parameters for managerial capability has moved from the team size to the size of work or the span of work.
Speaker: Yeah, I definitely yeah see the dual model we talked about the flattening of the structures as becoming very, very a key focal point now. And it's interesting. You're at an interesting size at LTM. You're all about $5 billion dollars revenue.
Speaker: So you've got a choice between you can go up into bigger greenfield opportunities and some might be traditional opportunities that they're disrupting.
Speaker: And you can also maybe go after some mid-market, some emerging stuff. How do you in your role make decisions around the cannibalization of the existing model and keeping the company on a growth trajectory and taking advantage of new economics. Is there some secret sauce that you're playing with here?
Speaker: Well, i there's no specific secret sauce, but you know there are and actually, if you go back and analyze it, it all comes back to the 101 of basic business. Firstly, fundamentally, i don't agree with the logic that, oh, there is a cannibalization happening with large clients.
Speaker: Let's go after the mid market or the small market size. For me, that's not the argument. you right so to To address the mid-market or the small and medium segment has to be a strategy on its own.
Speaker: right it has to Because there is a market potential, AI gives you a probably a lower entry barrier. so you know And you can convert some services, you can scale the services faster and help those customers to transform into the AI era is one argument.
Speaker: But with regard to the large customers where there is a cannibalization happening because there's a productivity headwinds, there are all the productivity benefits that we pass it on to the customer. My experience tells me feels that over the last one and a half year or so, wherever we have passed on the productivity benefit to customer, we've actually got a lot of work back in that. So, mean, we went through few quarters in specific customers where because of the productivity benefit that we passed,
Speaker: You know, we saw the revenue you know fluctuating a lot in some of the big accounts of ours. But now the same accounts, fast forward one year, we actually see a growth. So because the good thing that has happened is that the client, the the team which worked in delivering the productivity for our customers, ah but their skill level has actually gone one one notch above.
Speaker: Because they worked in some of the best technologies. They proved it to the customers that, you know what, I can do more work with less cost and so on. So that capability is there. And since that capability is there, we're getting more work coming our way.
Speaker: So for me, we have lot of work to be done doing with our existing customer. So I'm not getting distracted with at least going after the new segments. We have we have a good market segment already addressed.
Speaker: We have white spaces in our current definition of market segment, whether it's We can do a lot more in Europe, as an example. We can do a lot more in Asia-Pacific. Even in our identified verticals in North America, there is a still a lot of headroom to grow.
Speaker: So we want to be, you know, we want to remain focused on that and, you know, get the growth out there. So not stretching things too thin. Famous last words. No, it's just been a terrific hearing about how you're you know combating these market headwinds and continuing to grow the company. heart beneficial is it being part of a, you you're a part of a broader portfolio Alton and Torbro businesses. Do you do you benefit a lot from those those relationships?
Speaker: You know, firstly, Being part of the group, you know, which has engineering in its DNA makes all of us a bit, you know, all of our employees to, know, feel proud about being part of every engineering conglomerate.
Speaker: So that's one aspect of it. And the second thing is that there are a lot of domain expertise we tap into within L&T. right especially on the manufacturing side, you know on the defense side or on the you know engineering side.
Speaker: There's a lot of domain expertise. We tap into it so that we can take those expertise ah to our customers. and And third is, of course, know being part of the conglomerate, you you get access to certain client relationships, which which is you know very precious and you know huge for us for us to liberate. So I think we have benefited a lot being part of this ecosystem.
Speaker: Yeah, I mean, especially the manufacturing and engineering pieces of the business I can see doing do very well. yeah Well, look, Venu, it's been awesome having you share your thoughts again. And I know you're going to join us at our UK summit in November. Hopefully you'll get to meet any of our European HFS customers there as well. But really enjoyed having you as part of the conversation today. and And great to hear, you know, things going well in the in the new job. Well, it's not that new, but in the and the journey. withcraft After one quarter, you become veteran. So, I'm already done. Thank you, Phil. Thank you for inviting for this great conversation.
Speaker: Thank you for listening to From the Horse's Mouth. Don't forget to subscribe and like wherever you listen to podcasts. Got something to add to the discussion? Drop us a line at fromthehorsesmouth at hfsresearch.com or connect with Phil on LinkedIn.
