Education
MIT Ph.D. Computer Science. 2021 - 2025
- Advisors: Tommi Jaakkola, Regina Barzilay
- My PhD focused on developing multimodal diffusion/flow models for arbitrary state spaces. I used protein design as my test bed since protein data is multimodal and possess non-trivial structure. For instance, protein sequences do not have a natural left-to-right order but instead follows a Potts model that needs to be inferred from data.
- I collaborated closely with Nobel Laureate David Baker to apply my methods for scientific breakthroughs in designing novel proteins. This resulted in two Nature publications (one first author) and a first author Nature Methods publication.
Johns Hopkins University B.S. Computer Science and Applied Mathematics. 2014 - 2018
Experience
AI scientist, Xaira therapeutics. 2025 - now
- Protein structure prediction and design research. Developing novel machine learning techniques to learn from experimental data and design therapeutics with AI.
Research scientist intern, Microsoft Research. 2024
- Research in protein conformational sampling. Co-author on BioEmu.
Research advisor, Nvidia Fundamental GenAI Research. 2024 - 2025
- Research on scaling diffusion models for protein generation. Co-author on Proteina.
Research engineer, DeepMind. 2018 - 2021
- Research and engineering on protein folding: focus on data processing for large scale biological sequence data and investigaitng trade offs between graph- and attention-based neural networks for scaling to large proteins. Co-author on AlphaFold-multimer.
- Solely responsible for all the research engineering on DeepMind’s project of using deep learning to predict eye disease from medical images, published in Nature Medicine.
Software engineering intern, Instagram. 2017
Software engineering intern, Microsoft. 2016
Featured talks with recordings
Lecture 06 - Diffusion for Protein Generation.
MIT 6.S184: Introduction to Flow Matching and Diffusion Models.
Keynote: De novo design of protein structure and function with RFdiffusion (AI audience).
Neurips 2023 Workshop on Diffusion Models.
De novo design of protein structure and function with RFdiffusion (bio audience).
Neurips 2023 Machine Learning in Structural Biology Workshop.
Generative AI in drug discovery lectures.
MIT CSAIL
Atom level enzyme active site scaffolding using RFdiffusion2.
ML Protein Engineering Seminar Series.