Director, Robot Policy AI ML Job

Analog Devices’ Edge AI team is reimagining how machines perceive, decide, and act in the world around them. We’re developing intelligent systems that bring together best-in-class sensing with real-time AI. We're looking for AI/ML Engineers with expertise in learned robot policies, task planning, computer vision and decision-making under uncertainty. Your work will drive how autonomous systems reason about goals, decompose tasks, and adapt in dynamic environments. Required Travel: Yes, 10% of the time. Shift Type: 1st Shift/Days. Job Req Type: Experienced

Responsibilities

Prototype and deploy AI based robot policy architectures that operate under tight latency and compute constraints at the Edge; Develop task policies using techniques such as hierarchical planning, imitation learning, reinforcement learning or hybrid methods; Contribute to the evolution of our Edge AI stack by connecting planning with low-level controls and high-level semantic reasoning; Work with real-world and simulated sensor data to test and refine models; contribute to internal datasets and benchmarking tools; Participate in project planning, code reviews, and architectural discussions; take ownership of technical areas and deliver high-quality, reliable implementations; Stay current with trends in AI for robotics (e.g., expert and generalist policies, foundation models), and help integrate relevant advancements into our workflows

Requirements

3+ years of experience in AI, robotics, or computer vision; Master/PhD degree in a relevant field (e.g., Robotics, Computer Science, Electrical Engineering); Knowledgeable in robot locomotion, robot kinematics and robot dynamics; Proficient in Python, with practical experience using ROS, PyTorch or TensorFlow; Comfortable working with real-world sensor data (e.g., ToF, vision, IMU) and simulation tools like Gazebo, Isaac Sim, MuJoCo, OpenAI Gym or Unreal; Experience with DevOps/MLOps tools: Docker, CI/CD pipelines, version control (Git); Skilled at communicating technical insights, collaborating with multi-disciplinary teams, and contributing to shared architectural decisions

Nice to have

Experience integrating perception models with control loops, path planning, or robot operating systems (ROS2); Familiarity with industrial robotics, fixed arm robots, or autonomous systems in structured environments; Familiarity with teleoperation and/or learning from human demonstration; Contributions to open-source robotics or AI frameworks