About Me

My first exposure to ML was AlphaZero, in the midst of my senior year of high school. It left an indelible impression on me: mastery through self-play and inference-time search. Both immensely technical and efficient algorithms, yet still beautifully human at their core. I developed an interest in reinforcement learning, machine learning, and game theory.

I went on to study CS at Rice, where I worked on bioinformatics and founded the competitive Smash club Smash shaped how I think about problems: identify the outcomes, figure out how actions relate to them, and judge decisions on expected value, not just results.

After graduating, I attended UIUC for an MS focused on ML. My thesis work was on multimodal video understanding, fusing vision, language, and audio in a single transformer.

I came out of school right as LLMs were hitting the bleeding edge and started at Doowii, a ten-person education analytics startup where I built a NL→SQL system back when context windows were still 1024 tokens. How far we've come! Now I'm at Databricks, translating research into enterprise-scale products: work that lets me pull from my research background, engineering, and analytical instincts all at once.

Outside of work, I competed in the NeurIPS 2025 PokeAgent competition, applying actor-critic methods with Monte Carlo tree search to competitive Pokémon. A full-circle moment: AlphaZero's ideas applied to a game I've been deeply invested in as a competitor.

Finally, I spend a lot of time with philosophy. As intelligence becomes a cheap commodity, the questions that matter are increasingly human ones. What does it mean to create when AI can do it more efficiently — and in some ways (what ways?) more powerfully? How does AI shape the nature of humanity?