Machine Learning Engineer: Imitation and Reinforcement Learning for Robotics

About this role

Bedrock seeks a Machine Learning Engineer to develop behavior learning models for autonomous construction machinery. You'll design imitation and reinforcement learning systems, build robust data pipelines, and deploy models into real-world robotics applications across the country.

What you'll do

• Design, train, and validate behavior cloning and reinforcement learning models
• Build data ingestion, labeling, and management pipelines for high-quality training datasets
• Create evaluation metrics for assessing model performance across simulation and real-world environments
• Integrate ML models with simulation, systems, and infrastructure teams
• Deploy and debug models in production, handling latency and hardware constraints
• Collaborate with construction and engineering teams on physical-world problem-solving

What they're looking for

• Deep learning frameworks (PyTorch, TensorFlow, or JAX)
• Behavior cloning and reinforcement learning
• Production ML model deployment and maintenance
• Data pipeline engineering and management
• Performance metrics and evaluation methodologies
• Diffusion policies or Vision-Language-Action models (bonus)
• Robotics or autonomous systems experience
• Real-world debugging and system integration

Benefits

• $350M in funding backing the company
• Work on meaningful infrastructure and construction automation problems
• Collaborate with industry veterans from Waymo, Segment, and Uber Freight
• Flexible roles and locations (San Francisco or New York offices)
• Opportunity to transition cutting-edge AI from lab to real-world deployment