Experience
Building drug-discovery systems across ML, simulation, and molecular design
My recent work spans biotech startups and research-heavy teams, with a consistent focus on turning technically strong models and simulation methods into decisions scientists can use in practice.
4 industry roles ML + MD workflows Biotech product context Boston and global teams
Industry
Recent roles
- Built molecular representations and trained models for experimental ADME property prediction.
- Worked on de novo protein binder design and uncertainty-aware ranking workflows.
- Designed algorithmic library enumeration for optimized, synthesis-ready compound sets.
- Designed and implemented computational chemistry tooling to review molecules entering synthesis and assay workflows.
- Built an end-to-end MD + ABFE/RBFE platform in OpenMM with in-house Amber preparation.
- Engineered a high-throughput MD screening method that improved throughput by ~300×.
- Advanced internal AI and CompChem platforms to improve generated molecules and docking poses.
- Worked across AI/ML and medicinal chemistry teams to refine ADMET, active learning, and linker generation workflows.
- Led efforts within the PhysicsML team and contributed to the PhysicsML product direction.
- Helped deliver EXS-NNP, a GNN-based neural potential that outperformed ANI on coupled-cluster theory data.
- Evaluated ANI neural network potentials for small-molecule crystal polymorph prediction with DFT-like accuracy at force-field speed.
- Analyzed potential energy surfaces for 100 crystal systems and 500+ polymorph structures.
- Reported 98% DFT-to-ANI correlation in prediction quality.
Doctoral Research
PhD at UIC
- Implemented a generative deep learning model to guide conformational sampling of intrinsically disordered proteins and accelerate MD workflows.
- Developed a machine-learning enabled virtual screening pipeline against RPN11, a breast cancer drug target.
- Designed hybrid AI and molecular simulation workflows for SARS-CoV-2 main protease.
- Explored electron transport in molecular helices and other physics-rich modeling problems under Prof. Huan-Xiang Zhou.
Earlier Research
Foundation across chemistry and materials
- UIC research assistant work with Prof. Petr Kral on quantum chemistry methods including DFT, TDDFT, QM/MM, AIMD, and related analysis.
- MIPT work with Prof. Artem Oganov on USPEX-based prediction of stable and metastable europium nitride structures.
- IISc molecular dynamics simulations on zeolite MOFs and adsorption behavior of warfare-agent-like molecules.
- BARC and CSIR projects on porous carbon nitride catalysis and reaction isomerization pathways.
- ICT undergraduate work spanning dye prediction, cellulose chemistry, and silicene modeling.