I get excited when I get to build things for people. Things that make their lives >1% easier in some way or another.

That's the short version. 

Longer version, roughly backwards in time:

I recently got back to New Delhi after spending a year in San Francisco and 3 years in San Diego before that. Most of that time went into startups: working at an AI trading startup, and also, trying to build one of my own. We created products, shipped them, evolved or killed them, and started over. That loop is still running. I think you can either study product sense from the outside, or you can also learn it the way I did: ship things and observe what happens. Doing both is ideal.

I've done ML research and I've done B2B enterprise work. But what pulls me the most is consumer. I make a thing, I put it out there, and people either reach for it or they don't. This is the cleanest feedback loop. Of course with its own messiness, but this is the game I want to play. For the next decade.

Before startups, I was at Chainlink Labs, building a ranking system for their data providers using graph AI and Bayesian stats. Before that, during my master's in CS at UC San Diego, I worked at the ACES Lab w/ Intel on accelerating large-scale distributed AI inference. Both were real problems I had a ton of fun solving. But they were infra/plumbing problems. Sure, important ones, the kind on which great engineering systems depend, but I could never see who is on the other end, directly, deriving real value.

The work I'm most proud of is work where I can directly see who is enjoying the outcomes of my ideas and effort.

Early in my career, I spent ~2 years at a government AI lab in India building computer vision systems. Driver drowsiness detection for commercial vehicles. Face recognition that was shipped in millions of phones. This is the work people use everyday. They get tangible value. They have no idea who made it, their lives get better, and life goes on. This is the feeling I keep chasing.

The reason I ended up doing AI in the first place is more personal than I usually let on. In summer 2016, I almost fell asleep while driving. The crash was thankfully not fatal, but death's essence teased me. It rewired my sense of what is worth working on.

Anyway, after the government lab, I joined a research group at IIIT-Delhi where we published at ~10 top AI conferences: EMNLP, AAAI, NAACL, and others. At COLING, my work won an outstanding paper award. Over 1.5yrs, we built quant trading systems, decoded patterns in political speech (pre-LLM era), pushed on graph neural nets and RL. Through all this work, I got to collaborate with people from Tower Research, Bloomberg, Samsung, LG, NVIDIA, NUS, MIT, and more. Too much name-dropping but I can't help, I learned more from those geniuses than all of my degrees.

I run, lift weights. I make electronic music. I claude code.

In spare time, I think too much about why people do what they do. It might just be the non-technical version of the same problem I keep trying to solve.

I'm just trying to find my people, love the ones I cross paths with, and build fun things for humanity. Reach out to me if you want to chat. I'm always down.