About Me

Welcome to my website! I'm a PhD student in Robotics at MIT CSAIL, advised by Leslie Kaelbling and Tomás Lozano-Pérez.

I study embodied reasoning: how to connect high-level knowledge and planning with the messy realities of perception, control, and physical interaction. My goal is to build Lifelong Learning Robots that learn from data and still adapt to new tasks and environments, much like humans do!

My research bridges top-down, factored reasoning and planning with bottom-up, data-driven perception and control. Practically, that means designing and learning representations that capture the structure of a task (objects, goals, constraints) while learning the low-level skills needed to act in the world. I develop algorithms that let these layers talk to each other so a robot can plan abstractly, execute robustly, and reuse what it learns across contexts. Ultimately, I'm working toward general-purpose robot behavior that transfers across tasks, objects, and settings with strong sample efficiency and real-world generalization.