Summary
Alex Morehead is a Hopper Postdoctoral Fellow and affiliated researcher based in Berkeley who builds scientific foundation models and runs large-scale HPC deep learning experiments at LBNL and ICSI. With a Ph.D. in computer science from the University of Missouri, he has authored 20+ papers including ICLR and ISMB/ CASP16 contributions and created state-of-the-art protein representation and generative models for 3D molecules and protein–ligand docking. He developed the first SE(3) diffusion model for joint sequence-structure generation (MMDiff) during industry internships and led the Zatom family of models for scientific ML. Comfortable spanning academic rigor and production-ready code, he regularly contributes open-source tooling and reproducible experiment artifacts. Beyond modeling, Alex has taught neurodivergent students, worked in VC deal sourcing, and brings nine years of cross-disciplinary experience bridging ML, computational biology, and software engineering.
9 years of coding experience