It's my pleasure and privilege to be speaking to Srikanth Velamakanni of Fractal. Thank you so much, Srikanth, for taking time out. There could not be a more pertinent time to have this chat with you. We're seeing AI being leveraged as a soft power tool increasingly by world leaders. India is set to host the AI Action Summit later this year. OpenAI's Sam Altman was in Delhi just last month; you were there at the meeting as well. Srikanth, talk to us a bit about this global landscape that's evolving very quickly and what role AI plays in it, and also what is India's role now in the global landscape.
See, AI is one of the most game-changing general purpose technologies out there. It's on the same scale as the internet, mobile, cloud, but only much bigger. And it's probably bigger than all of those three things combined. And because it's general purpose, it means that it can be used across a wide range of applications in defense, in civilian uses, and a number of other places. And because it's such a powerful technology, it also has impact on how industries will be shaped, jobs will be shaped, and therefore whoever has control over this technology and the advances in this technology has higher bargaining power of power in the world. So you can see that people want to be ahead. And the DeepSeek revolution that happened just a month ago, right when DeepSeek was launched, before that the prevailing notion was that AI models are very expensive to build, not every country can afford to build them, and you know, people are so ahead of the race that don't compete with us. But what it has done is two things. Number one, it has told us that the AI revolution is just getting started. We are in the first chapter of this book, not the final chapter; there's a lot to do. And number two, that it's not that expensive. A country like India with 1.4 billion people absolutely must compete in this race. So that's why you're seeing increasing attention towards India, increasing attention globally towards who and how should AI models be built, and the excitement there as well.
Right. Srikanth, you spoke about how the cost of AI, the cost of deploying AI, is also going down. The Indian government in its India AI mission is also working towards that by providing compute capacity for accommodations for companies, for startups to sort of play around and basically develop that sandbox for Indian companies as well. What is Fractal looking at with respect to the India AI mission? There's multiple pillars there: the ethical pillar, the compute capacity pillar, the application layer. What sort of involvement do you have with the India AI mission?
I have been very fortunate to be a part of NASSCOM as a NASSCOM Vice Chair, as well as Fractal. We've had the opportunity to weigh in and support the India AI mission right from its inception, right from its conceptualization. So it's been a great experience. I think that the big shift today post-DeepSeek is about building foundation models for India and an Indian foundation model, not just restricted to Indian languages, but a foundation model that progresses us towards AGI. So if there's a way in which Fractal can contribute to that, absolutely we would love to explore that. Apart from that, there are other elements of the mission, like hardware capacity building, all those GPU capacity, some 7,000 crores being spent on that. Then investing in AI use cases or contributing to research in various universities towards greater AI. Certainly, as Fractal we can contribute to some of these, and also as NASSCOM, I think we have a good opportunity to be able to play an important role in how this mission achieves its objectives.
Right. Srikanth, now with AI coming in, India was the back office of the world when the IT wave came in. Now we're seeing a bunch of GCCs being set up. That's the next big buzzword. Every global big tech company now wants to do GCCs in India. How does India not limit itself to becoming a back office for AI? What do we have to do here? How do we change that mentality that okay, India is only for cheap talent? How do we change that?
So GCCs are Global Capability Centers. These are big companies thinking that India is the next destination for talent. Now this is not your old back office; in the old days of outsourcing, this is about doing the most cutting-edge innovation out of India. So these GCCs are a very credible, very important aspect of the company's overall mission, and my guess is that in the next 10 years, for every major company in the world, India will be the second largest destination of talent after their home country, if not the greatest destination of talent. So that is a power that we're talking about. Now coming to your question on AI and what role does India have to play on AI, you're absolutely right. India already has the world's best talent in AI. One in three AI researchers worldwide is of Indian origin. So therefore India is obviously contributing to the AI mission. Some of the greatest researchers I've met in AI globally have been of Indian origin. So what we can do differently in terms of India AI is to build these very cutting-edge models which might change the game and bring us closer towards artificial general intelligence, but also use AI to build cutting-edge applications for India and the world. Now India has 1.4 billion population. Almost every major tech product will be AI-infused. So Indian AI researchers can put AI to work and make the lives of 1.4 billion people better. And even from a government point of view, there are many, many use cases that can improve the citizens' experience of the nation. So these are the things that they can do as well. And the third thing is that multiple startups will get created. We'll have a bunch of AI entrepreneurs reimagining every business process or every problem and building companies from the ground up. So you'll see at the government level, at the industry level, at the startup level, and at the research level, Indian AI researchers and Indian AI entrepreneurs contributing.
Right. That's very interesting. Srikanth, also talk to us about government adoption of AI versus private adoption of AI. Are we seeing much faster adoption with respect to private companies now? Do you see world governments, not just India, but what's the scale that you also spoke about delivery of citizen services using AI? That is a government mandate. What will the vision by governments look like with respect to AI?
Yeah, so firstly let's look at the private sector. Why is the private sector quick to adopt? Because there's a clear motive to increase revenues, decrease cost, improve the customer experience. All of these things are very readily possible, and if you don't do it, you will fall behind. So companies that use AI are going to outperform companies that don't use AI. So it's very, very clear that companies have a very strong incentive to improve usage of AI and therefore improve their business prospects. Coming to the country, these things become a little less clear at that level. But if you think about government, it could be the biggest user of AI because government first of all handles enormous amounts of data sets. Then they're targeting a 1.4 billion population in India's case, which means that just a decision, any decision can be slightly improved; it could massively increase the experience. For example, education: what does India have as a problem? We don't have enough teachers in schools, and teachers are not well educated, and teachers may not show up. There's a teacher absenteeism problem. If you bring an AI to work along with teachers in schools, we can significantly improve education. India in the long term needs education and entrepreneurship. I believe that these are the two fundamental pillars that can take India and get us to the Viksit Bharat that we have been dreaming about. So education is one pillar that AI can influence dramatically. But another pillar which is equally important is that as India becomes larger and gets older, we need much more healthcare, much better quality of healthcare, and AI can significantly augment healthcare as well. Whether it is about creating AI radiologists so that every X-ray, every MRI and CT scan is read properly, or it is to make sure that every doctor is looking at all the symptoms and coming to the right diagnosis, or it is about making sure that healthcare outcomes in hospitals etc. are much better. In all of these situations, you'll see AI will make a big difference. And India is already investing behind that as well. From a government standpoint, India has a whole host of government hospitals and army hospitals and navy hospitals, so making sure that AI is infused in each of these is also an important use case. And then India has some schemes like direct benefit transfer and other kinds of schemes to make sure that the fraud, waste, and abuse come down to zero. Again, you can use AI to do that. Overall, every company is becoming a tech company, and every country should be a tech country as well. India should become more digitized as a country, and we have amazing accomplishments to speak about like UPI and other things. So bringing AI will make sure that India is digitally ready for Viksit Bharat.
