I'm a Member of Technical Staff on the ML team at Vinci, working on AI for physics simulation.

I came to this work through a PhD in computational physics at UC Santa Cruz, where I specialized in black hole dynamics and gravitational waves. Across 7 publications, 5 first-author in Physical Review D, JHEP, and JCAP, collaborations at ADMX and Fermilab, and end-to-end research pipelines, my work kept coming back to the same core question: where do models break down, and how do you build systems that tell you honestly when they are?

That meant building matched filtering pipelines across multi-terabyte datasets, evaluation systems that characterize failure modes across thousands of noise realizations, and simulation tools that push existing codebases into physical regimes they were never designed to handle.

Outside physics: I won top project at the Erdős Institute among ~200 PhD candidates — a 3D deep learning pipeline for MRI brain tumor segmentation from scratch. Dice = 0.83, 86% precision on held-out data.