ETH Zurich · EPFL
Robot Learning, Reinforcement Learning & Computer Vision
M.Sc. student in Robotics, Systems and Control at ETH Zurich, following a B.Sc. in Mechanical Engineering at EPFL and a study-abroad year at the University of Illinois Urbana-Champaign.
I am focusing on robot learning and computer vision: teaching robots to act from data through imitation and reinforcement learning, and to perceive their environment, with rigorous evaluation on situations they have never seen.
Available from September 2027. Open to conversations before then.
Zurich, Switzerland
Areas of focus
I am building toward robot learning and computer vision, on top of a foundation in control and mechanical engineering: learned behavior is only useful if it stays reliable on hardware and on unseen situations.
01
Imitation learning (behavior cloning, DAgger), reinforcement learning (PPO), policy architectures, and evaluation on unseen environments.
02
Visual perception for mobile robots: feature-based vision, camera geometry, and visual odometry, studied through Vision Algorithms for Mobile Robotics.
03
Robot dynamics, model predictive control, and optimal control, used both on their own and as strong baselines and experts for learning-based methods.
Featured project
A simulated two-wheeled robot must reach a goal while avoiding obstacles. I compared three ways of producing its behavior on the same 500 unseen test maps: a hand-written controller, a neural network that imitates it (behavior cloning + DAgger), and a network trained by reinforcement learning (PPO) that never sees the expert.
Projects
01Robotics · Machine Learning · Reinforcement Learning
A simulated two-wheeled robot learns to reach a goal while avoiding obstacles. I compared a hand-written controller, imitation learning (BC + DAgger) and reinforcement learning (PPO) on 500 unseen test maps.
1 / 4 Project 1 of 4: Learning-Based Robot Navigation
Background
Two separate tracks: academic education, and professional, teaching, and leadership experience.
Sep 2026 – Aug 2028 (expected)
M.Sc. Robotics, Systems and Control
Zurich, SwitzerlandCurrent graduate student
Aug 2025 – May 2026
Study-abroad year, Mechanical Engineering (Grainger College of Engineering)
Urbana-Champaign, Illinois
Credited as the third year of the EPFL B.Sc.
Sep 2023 – Jul 2026
B.Sc. Mechanical Engineering
Lausanne, Switzerland
Overall average 5.51 / 6.0
EPFL grading scale: 6.0 is the highest possible grade; 4.0 is the passing grade.
Sep 2024 – Jul 2025
Teaching Assistant — Physics, Thermodynamics & Computation
Teaching
Supported a cohort of 500+ students across four course assignments: Physics: Mechanics (Prof. P. Müllhaupt), Physics: Thermodynamics (Prof. S. Bréchet and Prof. J. Genoud), and Information, Computation, Communication (Prof. O. Lévêque). Led exercise sessions, graded midterms, and gave feedback.
Sep 2024 – Jul 2026
Ground Station Subsystem Engineer — Firehorn
Engineering
Mechanical design, prototyping, and testing of ground support components for competition-level rocketry; CAD, 3D printing, CNC machining, and cross-subsystem integration.
Sep 2024 – Jul 2025
First-Year Student Coach
Leadership
Mentored first-year students through their integration at EPFL, sharing study methods and strategies for the foundation year.
Jan 2021 – Jul 2023
Founder & Director
Entrepreneurship
Created and led an online academic support platform bringing together several thousand students, with revenue generated through advertising partnerships.
Capabilities
Tools and methods I work with across robot learning, control, engineering, and software.
Languages|French (native) · English (C1) · Spanish (B1)
Status
First semester of the M.Sc. Robotics, Systems and Control at ETH Zurich, with coursework in probabilistic AI, vision for mobile robots, and optimal control. Graduating in 2028.
Looking for internships in robot learning, reinforcement learning, and computer vision from September 2027.
Contact
Open to conversations and opportunities in robot learning, reinforcement learning, computer vision, and control.
Zurich, Switzerland