Research

Research Focus

I design novel machine learning architectures for atomistic simulation and deploy them on the problems that define modern drug discovery: crystal polymorph prediction, preformulation automation, and structure–property learning across scales.

The common thread is purpose-built method development—representations, model architectures, and inference frameworks shaped by physical constraints and validated on real pharmaceutical and materials science challenges. From committee-of-experts potentials to agentic LLM systems, each project translates foundational ML research into measurable impact on how molecules become medicines.