Build AI agents using decades of industry knowledge, TiE Bangalore's Venk Krishnan tells Indian IT
Venk Krishnan emphasises the importance of Indian IT firms utilising their extensive industry knowledge to develop AI applications. He warns that traditional entry-level jobs in coding and quality assurance may soon diminish. Instead, engineers mu...

Venk Krishnan, president, TiE Bangalore
Krishnan, who founded IT services firm NuWare more than 30 years ago and has backed startups such as Acko, bigbasket and Third Wave Coffee through NuVentures, believes India's biggest opportunity in AI lies in building applications and not in building its own large models.
"One of the things I keep saying is India is very good at applications. If you look at all our IT service companies, Infosys, Wipro, TCS, they have been supporting customers for 30-40 years in their mission-critical applications. And when they know how to do that, they know the domain," he told ET AI at the Bengaluru Tech Summit.
He wants the industry to put that experience to work in agentic AI. "Can we harness that? Can we use that domain knowledge to create agentic AI solutions or agents? One of the things which we need to really see is how can agents not just replace human beings, but can agents do better than the humans with their service. So I think we have a big opportunity in applications, in the application layer," Krishnan said.
While companies such as Sarvam are trying to build their own frontier models, Krishnan believes India should put its energy into applications instead.
"I think the game of frontier models is done and dusted, according to me, I may be wrong, but I feel that India should focus more on applications of AI and develop solutions for that," he said.
Engineers need to prepare for change
Krishnan warned that the entry-level work that has given jobs to large numbers of Indian engineers will not last long.
"Today, India is producing millions of engineers coming out of different colleges all over India. So where is this opportunity for them? I think things like coding, things like quality assurance, those are going to go away in maybe three to five years. So people have to be prepared. People have to know how to use AI, rather than just developing something," he said.
He said the Y2K era, when Indian IT firms learned their clients' businesses in detail, holds a lesson for the industry today.
"What happened with Y2K, I remember is our company learned the domain a lot, we learned their domain. Today, can we leverage that domain to create applications?" he said.
High cost of compute holds back startups
Krishnan said the high cost of running advanced AI models remains a major hurdle for Indian startups and he has spoken to the State government about the problem.
"We try to gravitate towards free models from China and others. But if you want to get real value or bang for the buck, you need to use some of these advanced models like Claude and OpenAI. So I think that costs money, that costs tokens. Can the government subsidise some of these?" he said.
He has also suggested a consortium that pools investment in compute and tokens for startups and he wants the government to give stronger support to a technology that he calls "bigger than the steam engine".
The Economic Times Business News App for the Latest News in Business, Sensex, Stock Market Updates & More.