Summary
Amirreza Razmjoo is an Applied RL Engineer based in Zurich with six years of experience translating advanced learning algorithms into robust robotic systems. He specializes in control for contact-rich and highly constrained scenarios—humanoids interacting with people and manipulation under uncertainty—combining model-based MPC, feasibility distributions, and generative policies (diffusion/flow matching) to boost success rates and reduce training burden. His research contributions include TT-PoE-MPC and CCDP, which have measurably improved task success and sample efficiency across manipulation and locomotion problems, and novel task representations like C-SDF and stiffness manifolds for safer, more stable robots. Comfortable spanning theory, physics-based simulation (Isaac Sim), and system-level integration, he has a track record of cutting training time by orders of magnitude while preserving real-world robustness. He is pursuing a PhD at EPFL and brings a pragmatic focus on deployable autonomy—making robots more intuitive and resilient in human environments.
5 years of coding experience
8 years of employment as a software developer
National Organization for Development of Exceptional Talents (Sampad)
Doctor of Philosophy - PhD, Doctor of Philosophy - PhD at EPFL
Bachelor of Science - BS, Mechanical Engineering, Bachelor of Science - BS, Mechanical Engineering at University of Tehran
Master's degree, Mechatronics, Robotics, and Automation Engineering, Master's degree, Mechatronics, Robotics, and Automation Engineering at Sharif University of Technology
English, Azerbaijani, Persian