About

I am currently a machine learning research scientist at Atomwise, where I focus on developing new machine learning techniques to aid drug discovery efforts. Prior to this, I completed my Ph.D. in theoretical chemistry at Harvard, where I was a joint student in the Coley and Shakhnovich research groups. My undergraduate training was in synthetic chemistry as part of the Knowles group at Princeton. I discovered early on in my graduate degree that I have two left hands, so I decided to leave lab work to the professionals.

I am broadly interested in the use of computational techinques (statistical, physical, or anything in between) to accelerate drug discovery with a particular emphasis on ultra-large search spaces. My graduate work initially focused on employing model guided optimization techinques to improve sample efficiency in high-throughput virtual screens (i.e., increase the rate at which we find molecular needles in the chemical haystack). A recurring theme of my research is an emphasis on cost because there’s no such thing as a free lunch. For example, while we might be able to improve a given technique’s accuracy by 10%, if subsequently double its cost, how useful is that improvement in practical terms? It’s non-trivial to answer that question, but it’s an important consideration to keep in mind.

Outside of research, I spend most of my time running or cooking. I ran my first marathon in grad school and just decided to keep going. I’m also into cycling, Formula One, and big dogs.