About

I'm from Barak Valley, Assam, studying Electronics & Communication Engineering at NSUT Delhi. Most of what I know technically I didn't learn in a classroom — I taught myself basic calculus around grades 6–8, and the physics intuition arrived before any formal study of it did.

I think of myself as a geometric thinker. Concepts click when I can see the shape of them first — the intuition lands, and the vocabulary just catches up. That's why I gravitate toward things that show why before how, and why I learn fastest by building something real instead of studying something abstract.

Right now that means Alin — a multi-agent personal AI I use daily, with an internal committee that argues a question out before it answers anything that matters. It also means showing up to hackathons and shipping working products on almost nothing: Mou, a civic-tech diagnostic built with Svar Chauhan, cost about $0.54 end to end and reached the global finals.

Longer term I'm pulled toward AI safety and governance — ideally working inside a frontier lab on the problems that decide how this technology actually gets built. Mechanistic interpretability is the path I'm circling, and I know there's real ground to cover before that's a door I can walk through.

Outside of code: cosmology and the parts of physics that sound like philosophy — spacetime, entropy, black holes, consciousness. Vedanta and Hindu philosophy. Urdu poetry. Layered, structurally ambitious fiction — Succession, Dark, Mirzapur, Shogun. And two Rottweilers, Zorah and Zairah, holding down Delhi and Guwahati respectively.