Description
The Multimodal Robotics Engineer designs and develops multimodal learning systems that combine camera, depth, tactile, IMU, language, and robot-state data into executable policies for dexterous robotic hands. The role also develops GPU-accelerated simulation environments and data pipelines, implements sim-to-real transfer, translates research questions into technical milestones, and supports engineers across machine learning, robotics, and hardware teams. Candidates need at least six years of computer science or comparable technical experience, including hands-on multimodal manipulation and simulation infrastructure work, along with Python, C++, ROS 2, and embedded or real-time systems expertise.
