I am a researcher and engineer with broad interests across statistical genetics, genomics, scientific computing, and more recently, applications of AI to scientific research.
I am currently a research consultant at OpenAI working on bio/ml applications, including developing evals.
Prior to that, I was a staff research engineer at Herasight working on statistical methods for polygenic embryo selection. Before that I led the research team at Gencove working on methods, product, and R&D applications of industrial-scale low-pass whole-genome sequencing, and prior to that, polygenic embryo selection again at Genomic Prediction.
My academic background is in physics and applied math.
Contact
Always happy to chat about AI/bio/ML; I can be reached on X or via email at h dot jeremy dot li at gmail dot com.
Recent writing
-
GeneBench-Pro: an updated, verified benchmark for AI agents in biology
Evaluating long-horizon scientific reasoning in genomics, quantitative biology, and translational biomedicine
-
GeneBench and current agentic capabilities in the life sciences
A new benchmark for evaluating real scientific workflows in genomics and quantitative biology