About

I'm a senior machine learning scientist at NewLimit, a biotech developing epigenetic reprogramming medicines — restoring youthful function to old cells by changing which genes they express, not their DNA. I build models that predict how a cell's state shifts under a perturbation, and use them to design better experiments. I helped build Ambrosia, the system we use to prioritize which transcription factor combinations are worth testing at the bench.

Before that I did a PhD at UC Santa Cruz, in Josh Stuart's lab, on interpretable machine learning for single-cell genomics. I developed VEGA, a variational autoencoder whose latent space maps onto pre-defined gene programs.