Maruan Al-Shedivat
Building @ Stealth AI Neolab • Previously: Genesis / CMU / Google / OpenAI • learn ⇄ imagine ⇆ build
Hey, thanks for stopping by! 👋
I was the Head of Machine Learning Research at Genesis Molecular AI, where I led the team and research behind Pearl, the first model to decisevely surpass AlphaFold 3 and currently the field’s best model for protein-ligand systems, deployed across multiple drug discovery programs at Genesis. My contributions also included inventing multiple methods and algorithms, such as a fast and accurate few-shot learning for binding-affinity prediction that has driven impact across all of lead optimization work done by Genesis internally and for partnerships.
As a scientist, I’m interested in building autonomous systems capable of solving problems at the intersection of AI x Science.
Previously, I completed my PhD in ML from Carnegie Mellon University, where I was advised by Eric Xing and worked on probabilistic approaches to multitask learning, supported by the CMLH Fellowship (2018/19) and Google PhD Fellowship (2019/21). I had also spent time at OpenAI (2017) and Google Research (2018/20).
In my spare time, I enjoy doing sports (mostly running and cycling), traveling, and going on hikes with friends and family. Also, I happened to grow up in Moscow and speak Russian.