Applied math & music at UC Berkeley. I build systems for deciding under uncertainty — a game-theoretic poker solver, a market backtester, and products that ship.
About
I'm a math person who likes building things that touch the real world. Most of what I make sits at the intersection of probability, code, and decisions made with incomplete information — which is also a decent description of both trading and improvising over a chord change.
When I'm not at a terminal I'm usually at a piano, reading about game theory, or losing a friendly game of poker for research purposes.
Experience
2025 — Present
UC Berkeley — Computational Biology Lab
Extracting signal from high-dimensional single-cell RNA-seq data: dimensionality reduction, clustering, and inference of cell–cell communication in neuroblastoma.
2024 — Present
Sylly
Built and shipped a full-stack academic planning app end to end — syllabus parsing, calendar sync, and an LLM study assistant.
Selected work
Computes approximate Nash-equilibrium strategies for an imperfect-information poker variant via CFR. Handles abstraction of the game tree, iterative regret matching, and average-strategy convergence — the same family of methods behind superhuman poker AI.
A backtester built for correctness first: point-in-time data to avoid lookahead and survivorship bias, modeled transaction costs and slippage, and realistic fill logic. Reports risk-adjusted performance rather than raw returns.
Signal extraction from high-dimensional, noisy biological data as part of a UC Berkeley research group: dimensionality reduction, clustering, and inference of cell–cell communication in neuroblastoma using Seurat and CellChat.
A shipped full-stack product: parses a syllabus into a structured plan, syncs deadlines to Google Calendar via OAuth, and runs an active-recall study chatbot grounded in retrieval-practice and spaced-repetition principles.
Background
Contact
The fastest way to reach me is email. I'm always happy to talk about markets, math, or anything I've built here.
Berkeley, CA