About the role
We are looking for an experienced Machine Learning Engineer to develop, train, evaluate, and deploy advanced ML models for real-world robotic systems. The role involves working across the complete machine learning lifecycle, including data engineering, model training, simulation, evaluation, and deployment on physical robots. The ideal candidate will have strong hands-on experience with PyTorch, distributed training, reinforcement learning, and production-grade ML systems, along with the ability to translate research into real-world robotic applications.
Responsibilities:
- Train and post-train robotic foundation models, including Vision-Language-Action (VLA) models, using imitation learning and reinforcement learning.
- Develop and optimize complex PyTorch training pipelines, including multi-GPU and multi-node distributed training.
- Build scalable training infrastructure, efficient data loaders, and reproducible experimentation pipelines.
- Develop simulation environments and train robotic policies that transfer effectively to real-world hardware.
- Design reward functions, task verifiers, and evaluation frameworks to measure model performance.
- Build data pipelines to convert raw robotic episodes into versioned, training-ready datasets.
- Deploy and optimize ML models on physical robots, ensuring reliable performance and efficient inference.
- Debug and troubleshoot issues across models, datasets, simulation environments, and robotic hardware.
- Own ML systems end-to-end, from problem definition and data preparation to training, evaluation, and deployment.
Requirements:
- 5+ years of hands-on experience training and deploying ML models in production or research environments.
- Strong expertise in Python and PyTorch, including writing and debugging complex training loops.
- Experience with distributed training frameworks such as PyTorch DDP, FSDP, DeepSpeed, or equivalent.
- Experience with multi-GPU or multi-node training and performance profiling.
- Strong understanding of machine learning fundamentals, linear algebra, probability, and optimization.
- Knowledge of reinforcement learning, imitation learning, and model evaluation.
- Strong data structures, algorithms, and systems fundamentals.
- Excellent software engineering practices and experience building maintainable, reliable ML systems.
- Demonstrated ownership of ML projects across the complete development and deployment lifecycle.
Preferred Skills:
- Experience with robotics, robotic foundation models, or Vision-Language-Action (VLA) models.
- Hands-on experience with simulation environments and sim-to-real transfer.
- Experience designing reward functions, task verifiers, and evaluation metrics for robotic tasks.
- Experience deploying and debugging ML models on real robotic hardware.