I am a research scientist at Google Brain. I am broadly interested in advancing science and intelligence, and where the ideas involve probability, programs, and/or neural nets.

I like to work simultaneously on fundamental research as well as systems to accelerate this research. In terms of systems, this includes Edward2 for specifying probability models as programs, Mesh TensorFlow for distributed computation, and Tensor2Tensor for deep learning research. Previously, I was a Ph.D. student at Columbia advised by David Blei and Andrew Gelman. I developed the original Edward language and was a member of the Stan development team.