I write software
and study optimization.

At Columbia, I study electrical engineering and work on Learning-to-Algorithm solvers for graph optimization. That means Transformer-based graph representations, reinforcement learning and curriculum learning, benchmarked on MaxCut.

Before Columbia, I led research at RPI on reinforcement learning for quantum circuit design and Ising models. I wrote three MDP formulations for ten quantum-circuit tasks, built a 190,000-instance synthetic dataset, and put together an open benchmark pipeline.

I also teach. In Fall 2026 I'm a Teaching Assistant II for ELEN E3801 Signals & Systems with Prof. Xiaodong Wang. At RPI I mentored more than 150 computer science students, and I've served as an IEEE special-track chair, a reviewer, and an organizer of the RLSolver Contest.

Outside research I build software: full-stack apps, backend services, asynchronous job pipelines and tools built on language models. The projects page has how each one works and what's still unfinished.