I build agents capable of navigating the web. My current research focuses on (1) grounding large language models (LLMs), by building custom web representations to better localize, and identify optimal actions, and (2) planning, by leveraging multi-agent methods to navigate the web step-by-step. Last year, I showed how to scale LLMs to learn from millions of workflows.
Previously, I have obtained my Master's of Science in Computer Science from Stanford University. And my Bachelor's of Science in Mathematics and Computer Science from McGill University.
I am fortunate to have worked with many wonderful collaborators and mentors. At Stanford, I improved the robustness of natural language technologies to dialectal variations under Diyi Yang. Prior to my graduate studies, I researched drug and protein language models to expedite the drug discovery process supervised by Jie Fu and Xue (Steve) Liu. I have also worked a lot on efficient machine learning. With Masoud Asgharian, I worked on an estimator for intrinsic dimension. Later, I developed numerical precision bounds for floating-point quantized neural networks with Vahid Partovi Nia.
In my free time, you can find me mastering the art of coffee brewing, refining my wine palate, enjoying sports with friends (🏃 🏓 🕺 🧗 ⛰), and capturing moments with my camera.