Machine Learning Engineer (Robotics/ Physical AI)

About the role:

• Build training data to operate on physical robotic platforms.
• Own the full machine learning lifecycle—from data collection and custom architecture design to edge deployment, simulation-to-real (sim-to-real) transfer, and continuous hardware optimization.
• Design, train, and evaluate deep learning, computer vision, and multimodal models (e.g., Vision-Language-Action or VLA systems) for robotic perception, reasoning, and manipulation.
• Apply synthetic data, domain randomization, imitation learning, and reinforcement learning within simulation environments before transferring policies to physical robots.
• Optimize, quantize, and compile ML models to run with low latency on edge compute hardware, sensors, and embedded robotic platforms.
• Build and maintain scalable data ingestion, labeling, automated testing, and MLOps/monitoring pipelines for real-world operational data.
• Work side-by-side with mechanical, electrical, controls, and software engineers to integrate AI models into safe, reliable physical machines.

Good to have:

• Strong ML fundamentals, computer vision, coding skills, and Claude Code usage
• Prior experience on perception models, multimodal datasets, and data pipelines for robotics.
• Strong experience in SLAM, computer vision, quick prototyping research papers, model optimization and deployment, Azure ML, claude code.
• Good understanding of SLAM, 3D reconstruction, real2sim, human exoskeletons and similar problems.
• Preference for hardware or full stack engineering; VLAs or Robotics learning.
• Prior 2-4 years of experience in Machine learning at high growth ventures.
• Advanced proficiency in Python and familiarity with Linux/Unix development environments.
• Practical background in computer vision, sensor fusion (cameras, LiDAR, IMUs), SLAM, or robot control/policies (imitation learning and reinforcement learning).
• Experience with edge inference optimization (ONNX, TensorRT), containerization (Docker, Kubernetes), Git, and CI/CD workflows.
• Bachelor's or Master's degree from Tier 1 Engineering College/ University in Computer Science, Robotics, or a related technical field.