‘It won’t be easy’: CureFit founder says building real AI agents needs more than clever prompts

Mukesh Bansal, founder of CureFit and Myntra, has questioned the gap between the rapid advancement of large language models (LLMs) and their limited real-world impact on business bottom lines.

In a post on X (formerly Twitter), Bansal wrote, “Everyone is building AI agents and yet so few AI agents in production doing real work. LLMs keep getting better and better and yet P\&Ls are not changing by even basis points! What’s going on?”

He stressed that building “production-class agents” requires a fundamental rethinking of business processes, not just automating existing workflows with prompts. “An agentic company is not like a traditional software or platform company at all. It is a cross between McKinsey and Infosys… maybe good for promotion but bad for business,” he quipped.

Bansal noted that enterprise-grade AI agents demand the best human talent working alongside LLMs to create solutions that can “dynamically evolve” and reshape industries. “If this was easy, agents would already be running the world! Revolutions are never supposed to be easy even if inevitable like this one is,” he added.

In a separate post, Bansal warned that India’s science and deep-tech ecosystem needs urgent strengthening in light of global geopolitical tensions and the coming AI wave. “This is no wake-up call. This is existential,” he wrote, calling for greater investment in curiosity-driven research, world-class infrastructure, and efforts to attract top Indian scientists back home.

“21st century demands nothing less. Viksit Bharat not possible without this,” Bansal emphasised.