AI-driven small molecule and protein design. Lead developer of GNN molecular property prediction platform. (PyTorch, DGL).Designed multi-task GNN molecular prediction algorithms and developed in-house library tool for GNN explainability.Mentored junior researchers in algorithm development for drug discovery and best practices for MLOps.Led projects for data collection and CRO management with the Medicinal Chemistry department for the development of our internal pipelines.

Modeled the behavior of micro-encapsulating particles through mathematical and computational/stochastic models developed by the team. Presented in 2019 APS Department of Fluid Dynamics conference. Work published APS Physical Review Letters E.

Researched applications of near-term short-depth quantum circuits (our current quantum computers) in machine learning under the supervision of Peter Shor and Ramis Movassagh.

Explored the construction of ancient solutions to the Curve Shortening Flow on the sphere, which obtain infinite length as the flow is run backwards, under the supervision of Prof. Christos Mantoulidis.

Created a (parallelizable) computer model for the inter-cellular interactions relevant to cancer formation utilizing stochastic numerical PDE methods to create a visualization of cellular membrane in cancerous protein interactions.Presented at the 2019 Joint Mathematics Meeting. Won the ‘Outstanding Poster’ award

In collaboration with Professors Luiz Faria and Carlos Perez-Arancibia generalized a numerical solver for 2-D PDEs to 3-D and compared its effectiveness to current state of the art algorithms for parabolic/elliptic BVP.

With the guidance of Professor Tobias Colding and graduate student Jiewon Park analyzed self-similar solutions to the Vortex Filament Equation. Categorized conditions for the existence of rotating solutions to the flow.