TCS CEO at DAVOS 2025: K Krithivasan Optimism for 2025, Trump’s Impact, and AI’s Future

Starting with the mood, we see early signs of revival, as we talked about in the last earnings call. It got validated in the last day's discussions we had. Particularly in the North American climes, they are reasonably upbeat with the potential for regulations to reduce and maybe some interest rate cuts if they happen. There seems to be some optimism, and we'll get to know as a new administration takes over and new policies come into play. But we see some optimism there. From a TCS perspective, as these technologies keep changing and evolving, the solutions we provide to customers will also keep changing, which means every organization has to be adaptable. We need to continue to train and retrain our workforce continuously to enable them to do different things. Similarly, the solutions we offer to our customers have to adapt, and your supply chain has to adapt continuously. You need to have resilience in your supply chain. That's what we mean by a continuously adaptable enterprise, and we help our customers achieve that while we try to become an adaptable enterprise ourselves.

My view is most corporates at a broad level know where these executive orders are going to be, so I am not expecting any shock there. As I said, they are going in with a position of optimism on this one. It's too early for them to take a decision, but many of them are going with the assumption that it will be very growth oriented.

I won't say there is a fatigue. Definitely, as we've been talking about in the earnings call, we believe that CY25 or FY26 would be a better year than the previous year. There are real signs that justify our optimism. We have seen our order book grow, and we've seen clients talking about new discretionary programs. So it's likely to be a better year. I don't think we have to worry about fatigue or further pessimism.

The strength of LLMs comes primarily from the ability to train them with a vast amount of data. That's where Indian IT companies would be very good. We'll be able to help our customers train these models, fine-tune those models, and new paradigms like agents come into play. How do you orchestrate this model? This is where our core strength comes from. Building a model – the reason I said we should not, we need not get into it – is that it may not give you a comparative advantage. There are already so many open source models available. You should be able to take some of them and train them. So there's no need to build your own model.