Now

I recently left Dhan after four years of building developer APIs, trading platforms, and agentic infrastructure. Right now I am between bets, and deliberately so.

Working on

• Researching the next bet — reading widely, mapping where new infrastructure is forming, and pressure-testing ideas before committing to one
• Writing actively — essays and working theses; An Operator's Dilemma is the first in a line of thinking about where interfaces go in an agentic world
• Building theses — on agentic AI, interfaces, and what the application layer looks like when anyone can build

Thinking about

• Whether the world converges to one chat window or a thousand hyper-personalised interfaces
• What the right primitives are for agentic commerce
• Long-tail bets that look small until they compound — same thesis, new domains