Experience

Senior AI/ML Scientist — Merck & Co., Inc.

Sep 2024 – Present · South San Francisco, CA
  • Built and implemented advanced generative AI models, including transformer and recurrent neural network (RNN) architectures with reinforcement learning (RL), to drive the inverse design of novel small molecules with targeted properties for structure-based and ligand-based drug design (SBDD/LBDD).
  • Developed state-of-the-art graph (GNN) and conformer ensemble-based (CNN) ADMET models, integrating uncertainty quantification to deliver predictive estimates and confidence metrics to medicinal chemists and accelerate design-make-test-analyze (DMTA) cycles.
  • Mentored interns and contributed to cross-team initiatives, including platform development and career outreach events.

Graduate Researcher — Caltech, Zhen-Gang Wang Group

Jan 2022 – Aug 2024 · Pasadena, CA
  • Developed a flexible workflow combining enhanced sampling molecular dynamics with unsupervised deep learning to elucidate phase diagrams for aqueous proteins and polymers; discovered key precipitation mechanisms and calculated quantitative structure-activity relationships.
  • Formulated generative AI diffusion models for polymer design and implemented a high-throughput free energy calculation pipeline to identify promising candidates for interfacial binding affinity.
  • Managed high-performance computing resources using Slurm and Ansible, optimizing resource allocation and increasing computational throughput for the research group by 10×.

Graduate Researcher — Caltech, John F. Brady Group

Sep 2019 – Dec 2021 · Pasadena, CA
  • Programmed CPU and GPU parallelized simulations of suspended bodies using C++, CUDA, and OpenMP, enabling computational fluid dynamics (CFD) study of self-propelled swimmers and collective phenomena.
  • Conducted theoretical modeling of hydrodynamic interactions for self-propelled bodies in Stokes and potential flow theory, unveiling symmetries across Reynolds number regimes and discovering relationships between body articulation and swimming efficiency.

Research Assistant — UC Berkeley, Kranthi K. Mandadapu Group

Aug 2017 – May 2019 · Berkeley, CA
  • Invented a dimensionless metric comparing lipid membrane out-of-plane bending with in-plane viscous forces, proving that viscous response can determine membrane stability and dynamics in biological systems.
  • Investigated phospholipid bilayer membrane dynamics using differential geometry and a unique balance law formulation to decipher coupling between surface geometry and in-plane flow behaviors.

Battery Cell Quality Assurance Intern — Nissan (Automotive Energy Supply Corporation)

May 2018 – Aug 2018 · Smyrna, TN
  • Designed and implemented an automated battery quality control system for daily testing of 25,000 cells, focusing on electrochemical stability, resulting in annual savings of $50,000.
  • Diagnosed electrode misalignment issues in manufactured battery electrode layers that contributed to cell failures and proposed tolerance revisions that cut $55,000 in annual material scrap costs.

Battery Management Systems Intern — Bosch

May 2017 – Aug 2017 · Palo Alto, CA
  • Evaluated electrode additive stability in lithium-ion batteries by analyzing dissolution under typical electric vehicle usage conditions to identify key degradation mechanisms and improve battery lifetime.
  • Initiated a comprehensive training program on plasma spectroscopy techniques and trained 3 colleagues, enabling the team to study trace lithium content in cycled battery components.

Research Assistant — Lawrence Berkeley National Laboratory, Nitash Balsara Group

Jun 2016 – May 2017 · Berkeley, CA
  • Determined electrochemical properties of solid-state polymer electrolyte batteries by performing electrochemical impedance spectroscopy across temperatures to identify electrode degradation mechanisms.
  • Fabricated lithium metal batteries with polymer electrolytes in a glovebox to study the effects of electrolyte molecular weight and cross-linking density on ionic conductivity to verify theoretical predictions.