I am a researcher, engineer, and entrepeneur working around artificial intelligence, interpretable machine learning, and computational biology.

I have a PhD in statistics from the University of Washington advised by Marina Meila, a BA in math from Columbia University, and participated in the W22 batch of Y Combinator with Uberduck.

At Uberduck I was responsible for development and deployment of AI pipelines for synthetic speech. I evaluated, trained, and tested new models, and helped create infrastructure for data, training, and inference. Here are some things we helped make.

My PhD research focused on sparse coding methods for interpretable unsupervised learning. We developed a mathematical framework for interpreting learned representations and applied sparse coding algorithms to estimate interpretations within this framework. This work was motivated by the need for interpretable reaction coordinates for enhanced sampling in molecular dynamics.

I have also published on both the neural connectome and hematopoiesis. For the former, we estimated cell-type specific connectivity in the brain from anterograde tracing experiments and decomposed these connectivities into interpretable components. For the latter, we explored clonal dynamics of hematopoiesis following autologous stem cell transplant using genetic barcoding.

Starting in August 2024 I will be an Applied Scientist with Amazon Connect!