AI Will Be A Trillion Dollar Industry By 2030: C Vijayakumar, HCLTech | Davos 2026

Yeah, first of all, thank you Shireen for taking time to talk to us. Of course we are very happy with the performance, but what we are truly excited is the opportunity that's ahead of us. While AI is a big inflection point in the industry, we've been very proactively looking at how to transform our services without worrying about any potential deflation or cannibalization of existing revenue streams. And I think that proactive approach has helped us really land very well with our clients in transforming the existing services. But what's even more exciting is a lot of completely new opportunities like in the tech vertical there is close to a trillion dollar of capex which some of it is already done, some of it is in the anvil. Now, every one of these big programs will need significant services in terms of planning out data centers, the professional services, the GPU infrastructure and the managed services around that. I think that's what we call as AI factory. That's a brilliant, very big opportunity. It could very easily be a billion dollar revenue stream for us in the coming years. And there are many such revenue streams that we've identified and it's still early days but we have some head start and that's why we're calling out our advanced AI revenue.

Yeah. So I started with talking about AI factory. I think it's a very small service line within the whatever 150 million which is about $600 million annualized revenue, it's probably less than 100 million. We think it can get to a billion dollar kind of revenue stream in 3 to 4 years. And similarly physical AI is again a very, very big theme. It's about while a lot of things have talked about agent take and AI in software development and operations, physical AI is all about AI's ability to understand, sense, understand and act with respect to the physical objects, whether it is warehouse management, port management, vision, all of that is what we call as physical AI, it's robotics plus AI infused workflows. I think that very well plays into our engineering heritage because a lot of these companies we've been working with them on their mechatronics and a lot of other areas, now it logically extends itself. I think physical AI is actually a trillion dollar industry by 2030, that's what the analysts talk about, but if I take the services, the places where we can play, that can also be another billion dollar opportunity. So I think we are just identifying these completely new areas. These are not where the traditional IT services we were constantly focused on. So while existing business will see some deflation, you can grow your wallet share and continue to offset that. But I think these new opportunities are really exciting.

I think we are thinking of at least in the next two to three years we should get to two to two and a half billion in size on this completely new service lines. These were not existing, this is not about modernizing existing services through AI. These are completely new discretionary spend areas.

It's an aspiration but of course we have to see.

I think it's still evolving and there is no one silver bullet as to how do you price it, but at a conceptual level moving all the work to outcome based engagements, that's really where the value is. And instead of really looking at time and material or even output based, can we focus on an outcome? Like a simple thing like if you're really focused on a business operation where employee claim settlement, right? For example, somebody puts in a claim, it gets reimbursed after a week or so. But everything can be completely automated: the receipt can be read automatically, it can figure out it is for food or hotel or any other expense, put it in the system, validate it and the payment can be done the same day or next day. So you're really focusing on end to end value chain and focused on an outcome. If you can commit to a client that you will deliver all their claims within the same day or next day, then I think that takes out so much of clutter in the entire system. So I think it's more outcome based constructs which are...