Member of Technical Staff, Robot Learning

What you'll do

• Develop and train end-to-end robot learning models for manipulation, navigation, and human-robot collaboration tasks.
• Build and optimize data pipelines for collecting, preprocessing, and augmenting teleoperation demonstration data.
• Train machine learning models on GPU clusters using distributed training techniques and hyperparameter optimization.
• Deploy trained models to physical robots, debug sim-to-real transfer issues, and iterate based on real-world feedback.
• Implement evaluation frameworks with standardized benchmarks and generalization testing across various objects and environments.

What we're looking for

• Extensive experience in end-to-end robot learning, including imitation learning, reinforcement learning, and vision-language-action models.
• Proficiency in Python with deep expertise in PyTorch, JAX, or TensorFlow for training and deploying deep learning models.
• Hands-on experience building and maintaining data pipelines for robot learning, including teleoperation data collection and preprocessing.
• Expertise in deploying learned policies on physical robots, debugging sim-to-real transfer issues, and evaluating model performance.
• Familiarity with robotics middleware (ROS/ROS2), control interfaces, sensor processing, and state-of-the-art research in robot learning.
• Experience with distributed training on GPU clusters using Azure Machine Learning, Kubernetes, or similar infrastructure.