I’m a machine learning data scientist passionate about building new ML-backed systems and applications that delight their users. My primary interests are MLOps and scientific devops in general. I spent a decade studying dark energy and dark matter as an academic researcher, and another decade modeling product demand, ad engagement, lifetime value, and recommendations at two startups, Netflix, and now Discord. I’ve also occasionally served as a pro bono Data Ambassador for DataKind.

I hold a PhD from Harvard and a BA from Berkeley with specialties in astrophysics and cosmology, and I held research positions at Caltech and the University of Chicago. I’ve made my publications from this period available under Publications.