I am an AI Engineer at RIVR Technologies AG, where I work on imitation learning and reinforcement learning to equip delivery robots with manipulation capabilities, leveraging a data flywheel approach. Before joining RIVR, I obtained my PhD from École Polytechnique Fédérale de Lausanne (EPFL) while working as a research assistant in the Robot Learning and Interaction (RLI) Group at the Idiap Research Institute, supervised by Dr. Sylvain Calinon.
My research focuses on controlling robots in highly constrained scenarios, particularly loco-manipulation in contact-rich environments. I am interested in developing adaptive, efficient, and intelligent robotic control methods that enable robots to interact with the physical world more naturally and safely. To achieve this, I explore Generative AI (e.g., diffusion models, flow matching), control techniques (e.g., MPC, RL), model composition (e.g., product of experts), system dynamics, control theory, and physics-based simulation (e.g., Isaac Sim).
More broadly, I am interested in bridging the gap between theory and real-world robotic applications, leveraging AI-driven models to improve robot autonomy and interaction. My goal is to push the boundaries of robotic learning and control, making robots more intuitive, adaptable, and physically intelligent.
News
(2026/2) I joined RIVR Technologies AG as an AI Engineer.
(2026/1) Our paper Geometry-aware Policy Imitation is accpeted to ICLR.
(2025/6) Two papers got accepted to IROS: 1. CCDP: Composition of Conditional Diffusion Policies with Guided Sampling, and 2. A Smooth Analytical Formulation of Collision Detection and Rigid Body Dynamics With Contact.
(2025/5) Our paper entitled Robust Contact-rich Manipulation through Implicit Motor Adaptation is accepted to IJRR.
(2024/10) I will be joining Honda Research Institute Europe for a six-month internship, working under the supervision of Dr. Michael Gienger and Dr. Fan Zhang.
(2024/09) Our paper Robust Manipulation Primitive Learning via Domain Contraction is accepted to Conference on Robot Learning (CoRL), 2024!
(2024/07) Our paper Configuration Space Distance FIelds for Manipulation Planning. has been nominated for the outstanding paper award at RSS 2024!
(2024/06) Our paper Logic-Geometric Planning and Control Using Graph of Tensor Networks is accepted to RSS24 workshop: Frontiers of optimization for robotics !
(2024/06) Our paper Logic Learning from Demonstrations for Multi-step Manipulation Tasks in Dynamic Environments is accepted to RA-L!
(2024/05) Two papers are accepted to RSS 2024: 1- Logic-Skill Programming: An Optimization-based Approach to Sequential Skill Planning, and 2- Configuration Space Distance FIelds for Manipulation Planning.
(2024/01) Our paper D-LGP: Dynamic Logic-Geometric Program for Combined Task and Motion Planning is accepted to ICRA 2024!
(2024/01) Our paper Learning Joint Space Reference Manifold for Reliable Physical Assistance is accepted to IROS 2023!