I'm currently Research Scientist Intern @ Google Robotics and PhD candidate in robotics and optimization @ Stanford University working with Zachary Manchester and Mac Schwager.
I am passionate about developing fast optimization algorithms for robotics applications. I am designing and implementing a differentiable physics simulator and leveraging this tool for model predictive control, trajectory optimization and reinforcement learning tasks in robotic locomotion and manipulation. Previously, I implemented optimization algorithms enabling game-theoretic reasoning for autonomous vehicles
Previously:
- research scientist intern at Google Brain Robotics
- software engineering intern at Aurora Innovation
- undergraduate student at Ecole Centrale Paris
- Dojo: A Differentiable Physics Engine for Robotics
- Fast Contact-Implicit Model-Predictive Control
- ALGAMES: a fast augmented Lagrangian solver for constrained dynamic games code
| 🎁 project name | 📖 brief description | language |
|---|---|---|
| Silico.jl | a unified formulation for collision detection and contact simulation |  |
| Dojo.jl | a differentiable physics engine for robotics | |
| ContactImplicitMPC.jl | a model preditive controller for robots that make and break contact | |
| Algames.jl | SotA dynamic games solver | |
| Unitree A1 | Boston Dynamics Atlas | Panda Arm | Tugging Drone |
|---|---|---|---|
 |
 |
 |
 |
I'm using Robin Deits' MeshCat.jl and developing RobotVisualizer.jl to make it prettier.