

Machine Learning Research Scientist
Training Foundation Models to Decode the Complexity of Life
About Me
I am a machine learning research scientist at EvolutionaryScale. I work at the intersection of machine learning and computational biology. In a nutshell, I train foundation models—from protein language models to diffusion-based architectures—to understand biological systems and design functional proteins for real-world applications in drug discovery and structural biology.
Previously, I was a Postdoctoral Fellow at Mila and McGill University in Montréal, where I worked with Prof. Adam Oberman and Prof. Ioannis Mitliagkas. My research during this time focused on generative modeling, distribution shifts, and optimal transport. I started applying some of these methods on biological applications like single-cell trajectory inference and protein design. I also co-created the open-source TorchCFM package to share our work on Flow Matching models.
I completed my PhD at INRIA Rennes in Brittany, France, under the supervision of Prof. Nicolas Courty and Prof. Rémi Flamary. My work explored the intersection of optimal transport and deep learning, with applications to domain adaptation, learning with noisy label, and generative modeling. Another core element of my PhD was the study of minibatch optimal transport. My thesis defense and manuscript are available online.
My résumé can be found here (last updated: April 2025). Outside of research, I enjoy exploring New York City with my wife and love getting outdoors—whether it’s hiking in the Hudson Valley or scuba diving in the Caribbean.
Research Interests
For a complete list of my publications, visit my Google Scholar. My current areas of focus include:
- Next-Generation Generative Models
- Designing novel diffusion models and flow matching techniques for protein modeling, computer vision, and tabular data.
- Advancing discrete generative modeling through masked language models, discrete diffusion, and discrete flow matching.
- Protein Structure and Sequence Generation
- Building deep generative frameworks for 3D protein backbone generation and sequence optimization.
- Modeling molecular evolution and generating functional proteins through learned priors and structure-aware objectives.
- Engineering protein-protein interactions, binder design, and multimer assembly via spatial constraints, evolutionary signals, and machine learning-based docking.
Feel free to reach out if you’re interested in discussing any of these research areas!