I build the things I wish existed.

I've been building software products since 2013 — first inside other people's startups, now mostly through companies of my own. The pattern has never changed: learn by shipping.

The early years were employment and contracts — Paytm (seller platform, then growth engineering), four years building a fintech app at Pei that Americans used every day, a stint at Flex after the acquihire. Before all of that there was Unstudious, my first venture, an EdTech idea good enough that Matrix Partners and Tracxn came asking. It didn't become a company, but it made me one.

In 2019 I started Droidsize, the product studio where most of my work lives today — SaaS and AI products like Domain Collective, Sparkles, and TripleWave, built on shared rails that make every next product faster. In 2023 my wife and I founded Charge23 Labs to build Chargespot, EV charging for India — DPIIT-recognised, incubated at AIC Sangam, and the hardest problem I've picked yet. On my own time I build in the Polkadot ecosystem: Relaycode is funded by the Web3 Foundation, with its second milestone delivered in early 2026.

Since late 2025 I've built almost everything with AI agents in the loop. It's the biggest shift in how I work since I learned to program — I ship faster today than at any point in thirteen years.

What I'm optimising for now: fewer, better products. Public work over private drafts. And the kind of compounding that only shows up when you keep building for a long time.