As an ML Engineer, Manipulation, you will develop and deploy learning-based manipulation systems that enable mobile robots to interact reliably with the physical world in dynamic human environments. You'll build perception-to-action models, training datasets, evaluation tooling, and deployment pipelines that improve robustness, generalization, and safety for real-world manipulation tasks at scale. Your work will directly impact the robot's ability to perform complex interactions consistently across real sites with minimal special-case engineering. Responsibilities • Develop learning-based manipulation models for end to end sensor-driven interaction (e.g., reaching, motion generation, and execution in dynamic environments). • Build and maintain manipulation training pipelines: dataset creation from robot logs/teleop, action representations, augmentation, and distributed training. • Design evaluation metrics and regression tests that quantify manipulation reliability, recovery behavior, and safety in real environments. • Develop sim-to-real workflows for manipulation learning, including simulation environments, domain randomization, and failure-mode testing. • Optimize and distill models for edge deployment; benchmark latency, memory use, and stability on target hardware. • Partner with the AI platform team to integrate policies with control and safety systems, and validate end-to-end performance on robots. • Analyze field performance, identify dominant failure modes, and drive iterative improvements through data collection and targeted retraining. Basic Qualifications • Bachelor's or Master's degree in Robotics, Computer Science, Electrical Engineering, or related field (PhD a plus). • 3+ years of experience applying ML to robotics manipulation, visuomotor control, or sequential to sequence models. • Strong proficiency in PyTorch and experience building reliable training/evaluation pipelines. • Strong software engineering skills in Python; ability to collaborate across ML and robotics teams. Preferred Qualifications • Experience with Vision-Language-Action (VLA) models, behavior cloning, and/or transformer/diffusion policies for robotic control. • Experience with sim-to-real training for manipulation (Isaac Sim/Mujoco or similar), including domain randomization and synthetic data. • Experience deploying ML models to edge hardware (ONNX/TensorRT, quantization, performance profiling). • Familiarity with safety-critical robotics integration and designing fallback/recovery behaviors.