Plotting India’s position in the AI race based on the five-layer AI stack, Nasscom Chairman Srikanth Velamakanni lists private capital and the need for a risk-appetite approach as the major bottlenecks for India’s progress. Acknowledging India’s efforts to punch above its weight on AI applications, Velamakanni voices his optimism in India’s AI future as an innovator economy in an exclusive interaction with businessline.
Do you feel that India is doing enough in terms of its AI initiatives?Jensen Huang of NVIDIA has been talking about the five-layer AI stack and how it starts with electricity and energy. On that India is doing a phenomenal job but for chip infrastructure, we don’t have companies like NVIDIA, AMD, ASML. That’s the reality. So we need to work more on this layer and collaborate while we build our capacity. We also don’t have a frontier model like that of OpenAI or Anthropic. If you look at purpose-built models, we have Sarvam for voice, Fractal’s Vaidya for healthcare. On applications, there are more than 3,500 deep tech companies in India, which are all AI companies right now. There’s an incredible amount of activity happening on building AI apps. So, the biggest weaknesses for India today stems from the chip layer and model layer. We have to do more.
