Second year CS at BITS Pilani, graduating 2029. Most of my time goes into two questions I can't shake: how machines actually learn, and how markets actually move. I don't get very far reading about either one, so I build things instead and see what falls over.

At 18 I built and shipped a product on my own, front to back, with nobody to hand the hard parts to. It grew entirely on its own with no marketing behind it. That taught me more about what people actually need than any amount of planning would have.

These days I do venture research at SH1P, which mostly means reading a lot of early stage pitches and trying to work out which ones are real.

Learning  ·  DSA and competitive programming in C++. Slow, unglamorous, worth it.

Building  ·  A backtesting engine for trading strategies, and a CNN that classifies music genres from spectrograms.

Chasing  ·  A better sense of what separates a real idea from one that only sounds like one.

AntarYatra  ·  A mental health platform I designed, built, deployed and grew by myself. No team, no budget, no marketing spend. Everything it got, it got organically. _{React · Node · PostgreSQL · Python}

I think for a long time before I start. Longer than I should, probably. Once I do start, I learn by messing with things rather than reading documentation front to back, which means I break a lot and then find out why.

I work best on my own. That's a preference and also a limitation, and I know which one it is. I'm slowly getting better at building things with other people instead of around them.

The thing that annoys me most is people who talk about work instead of doing it.

Whether you can tell a good founder from a lucky one before the outcome is known.

Why some models generalise and structurally similar ones just memorise.

Whether thinking carefully and shipping quickly are actually in tension, or whether I've just been telling myself they are.

If you've got a problem that's actually interesting, or you think I'm wrong about something above, my inbox is open. I answer people who have clearly thought about what they're writing.