Robotics Research & Engineering Intern at Stealth Startup

As a Robotics Research & Engineering Intern at Stealth Startup, you will bridge the gap between learned policies and physical hardware. You’ll be responsible for training reinforcement learning models, characterizing robot arm performance, and conducting systematic experiments. This hands-on role involves working across AI, controls, and mechanical engineering to solve complex manipulation challenges in real-world environments.

• Develop and deploy RL and imitation-learning policies on physical hardware, specifically focusing on reward design and investigating policy failure modes.
• Calibrate and characterize robot arms and grippers to map operating ranges and build repeatable manipulation routines for delicate or compliant objects.
• Design and execute rigorous physical experiments to evaluate system reliability, documenting every variable and failure to guide future hardware and software iterations.

• Deep practical experience training RL or imitation-learning policies, including a strong grasp of reward design, training stability, and sim-to-real transfer challenges.
• A systematic, experimental mindset with the ability to troubleshoot hardware, isolate variables, and explain technical failures clearly to a multidisciplinary team.
• Strong mechanical intuition and a foundational understanding of how grasping, compliance, and force feedback influence complex robot manipulation tasks.

===

• Work at the bleeding edge of robotics by tackling the “force and contact” problem, moving beyond simple automation into complex, unpredictable physical interactions.
• Gain end-to-end ownership of the experimental loop, seeing your reinforcement learning and imitation-learning policies go from training scripts to real-world robot arm execution.
• Direct mentorship and collaboration in a high-intensity startup environment where you help decide the next generation of hardware and software tests.