Summary
Peng Xu is a machine learning engineer with nine years of experience building intelligent autonomous systems, currently working on Integrity AI and foundational models at Meta. He has a strong robotics and decision-making background from roles at DiDi, Amazon, and Honda Research Institute, specializing in behavior planning, motion planning, and handling occlusions in autonomous driving. Peng's career blends research and production delivery—designing planning/control algorithms, simulating in CARLA, and deploying robotic behaviors for delivery robots. Early work at Huawei gave him hands-on systems experience from Android display modules to ROS-based robotic applications and activity recognition with TensorFlow. Based in Menlo Park, he pairs an MS in Electrical Engineering with a prior master's in Naval Architecture, an uncommon cross-domain foundation that informs his systems-level thinking. Known for making autonomous agents behave more human and robust in dense, occluded environments, he moves seamlessly between research prototypes and scalable production systems.
9 years of coding experience