Machine Learning · Software Engineering
I build ML systems for scientific, transportation, and real-world data — from plasma forecasting to spatial risk mapping.
Featured Work
Autoregressive forecasting of tokamak plasma density across 1,100+ real KSTAR shots.
Two-stage spatial ML for high-risk intersection prediction, paired with retrieval-grounded safety explanations.
Testing retrieval and prompting strategies for reliable LLM evaluation of K–12 science classroom dialogue.
Experience
About / Skills
I'm interested in building machine learning systems that are not only accurate, but also understandable, testable, and useful in real-world settings. I especially enjoy the process of figuring out why a model behaves the way it does, identifying where it breaks down, and iterating on it through careful experimentation. My recent work has taken me across scientific forecasting, transportation safety, and LLM evaluation, and I enjoy working on problems that combine technical depth with practical impact.
B.S. Engineering Mathematics & Statistics, Minor in EECS — UC Berkeley, Aug 2023 – Expected May 2027