(Senior) AI Engineer - Reinforcement Learning Manipulation

In this role you help push autonomous dexterous manipulation through reinforcement learning at RIVR, an Amazon-backed robotics company. You will work across simulated and real-world data to tackle contact-rich manipulation challenges in dynamic environments. You’ll translate multimodal sensor input into precise motor commands, shaping capabilities for last-mile delivery robots. This position blends cutting-edge AI with practical robotics to deliver scalable, real-world impact.

• Develop reinforcement learning algorithms for contact-rich dexterous manipulation
• Integrate vision, depth, tactile, and proprioceptive data into control policies
• Design and test solutions for real-world manipulation tasks (e.g., hand-offs, door handles)
• Collaborate with the foundation model team to leverage simulated and real-world data

• Strong background in robotic manipulation, dynamics, grasp synthesis, trajectory optimization
• PhD experience is advantageous but not required with equivalent research experience
• Deep learning fundamentals (supervised, self-supervised) and reinforcement learning (MDPs, policy optimization, model-based vs model-free, exploration-exploitation)
• Experience deploying neural networks on hardware; production-level C++ and prototyping in Python

• collaboration with cross-functional teams
• strong communication of complex ideas
• problem solving in uncertain, real-world settings
• reinforcement learning
• deep learning
• robotic manipulation