Senior Physical AI Engineer

About this role

Samsung SDS America AI Team is researching the next generation of intelligent robotic systems that operate in the real world. Working with multiple robot hardware partners, we use their platforms to develop complete embodied AI systems, spanning data collection through teleoperation, policy training, and deployment on physical robots in real-world environments. We are looking for a Senior Physical AI Engineer to join the team developing this end-to-end pipeline. You will work hands-on with humanoid robots, collecting demonstration data through teleoperation, training learning-based policies, deploying them on the physical platform, and iterating through live test runs in real environments, not just simulation. You will also research dexterous manipulation, advancing the hand and grasping capabilities that enable robots to handle complex tasks. This role sits at the intersection of robotics, embodied AI, machine learning, and production software engineering. You will connect high-level AI models to physical robot behavior, enabling robots to perceive, learn, adapt, and act safely in dynamic environments. What sets our lab apart is the shared ownership and the path to impact. You will work alongside a small team that takes real ownership of what we develop here, helping shape research directions, building systems directly, and seeing your work move quickly from ideas to robot experiments on real hardware. Samsung SDS builds and operates systems that support manufacturing at scale across thousands of factory lines worldwide. The Physical AI capabilities we research here are aimed at that production environment, with a path that extends from a single robot today toward coordinated multi-robot systems on the factory floor. This is a hands-on research and engineering role spanning teleoperation, policy training, real-world deployment, low-latency control systems, simulation, and scalable AI training pipelines. If you enjoy working with robots that move, learn, and interact with the physical world, this role is for you. Responsibilities Design and develop intelligent robotic systems powered by modern AI and machine learning techniques Build scalable training pipelines for embodied AI and robot learning applications Develop production-grade robotics software in Python and related frameworks Train and deploy learning-based robotic policies using reinforcement learning and imitation learning approaches Integrate AI models with physical robot actuators, camera systems, and sensor pipelines Implement robotics control algorithms, kinematics, dynamics, and sensor fusion systems Work with simulation and synthetic data generation environments including Isaac Sim, MuJoCo, Sapien, or similar platforms Optimize low-latency inference and control systems for real-world robotic deployment Collaborate across robotics, AI, systems, and hardware teams to bring intelligent robotic platforms into production Contribute to the architecture of next-generation embodied AI systems Requirements 5+ years of industry experience in robotics, embodied AI, physical AI, or autonomous systems Bachelor's degree in Computer Science, Artificial Intelligence or relevant field Strong Python programming skills with experience building scalable production systems Hands-on experience building, training, and deploying machine learning systems for robotics applications Deep familiarity with reinforcement learning, imitation learning, and modern ML frameworks such as PyTorch Strong understanding of robotics fundamentals, including: - robot control systems - kinematics and dynamics - perception and sensor fusion - camera and tactile sensor integration Experience with robotics middleware such as ROS or ROS2 Experience working with simulation environments such as NVIDIA Isaac Sim, MuJoCo, Sapien, or equivalent platforms Strong debugging and systems engineering skills across software and hardware boundaries Experience deploying robotics systems into real-world physical environments Preferred Qualifications Experience with humanoid robotics or manipulation systems Background in autonomous systems or embodied foundation models Experience with GPU acceleration and distributed training systems Familiarity with synthetic data generation and sim-to-real transfer techniques Experience optimizing robotics systems for edge or embedded deployment Benefits Why Join US Get hands-on with a real humanoid robot. Train it, test it, and iterate on real hardware, not just in simulation. Research dexterous manipulation, pushing the frontier of robotic hands and grasping for complex real-world tasks. Build Physical AI that deploys at scale, into real manufacturing environments running across many thousands of factory lines worldwide. Work on frontier Physical AI and robotics technologies that bridge advanced AI with real-world robotic behavior. Collaborate with a highly technical, fast-moving research/engineering team. Help shape a new robotics effort from its earliest stage. Samsung SDSA offers a competitive salary, and a comprehensive suite of programs to support our employees: Top-notch medical, dental, vision and prescription fully covered for You and Your family Wellness program Parental leave 401K match and savings plan Flexible spending accounts Life insurance Paid Holidays Paid Time off Additional perks The base salary range for this Mountain View, CA role is $200,000 to $270,000. Individual pay is determined by multiple factors, including job-related skills, experience, relevant training, and specific geographic location. Your recruiter can share more about the total compensation package—which may include performance bonuses, and comprehensive benefits—during the hiring process. Samsung SDS America, Inc. is an equal opportunity employer. All qualified applicants will receive consideration for employment without regard to age, race, color, religion, sex, sexual orientation, gender identity or expression, national origin, disability, status as a protecte