Robotics ML Engineer — Real-World Autonomy

Rainfall Ventures; San Francisco (CA); On-site; USD 150,000 - 200,000; Full time

Benefits offered by this job

Conference travel covered

Medical, dental & vision plans

Meals stipend

Job summary

Rainfall Ventures in San Francisco seeks a Machine Learning Engineer to design and advance a modular autonomous robot stack that fuses Vision-Language-Action models with purpose-built modules for grasping in unstructured environments.

You will implement action refinement and safety layers, architect interfaces for model swapping, and build robust data pipelines and high-speed inference software for production deployment.

Qualifications

• PhD or MS in Computer Science, Machine Learning, Robotics, or equivalent technical discipline.
• Deep expertise in ML fundamentals and reinforcement learning; familiarity with ML frameworks (PyTorch, TensorFlow, Ray).
• 3+ years of proven track record deploying ML systems from research to production.
• Hands-on experience with model lifecycle management: training, deployment, and maintenance.

Responsibilities

• Design and develop a modular robot autonomy stack that combines Vision-Language-Action models with specialized modules for grasping and dexterous behaviors in unstructured environments.
• Implement action refinement and safety layers to post-process VLA outputs: constraint satisfaction, collision and force guards, smoothing, and runtime monitors for safety-critical deployment.
• Architect clean interfaces around base VLA models to enable swapping, benchmarking, and upgrades as SOTA evolves.
• Design and maintain data collection and curation pipelines for production robot fleets.
• Build high-speed robot autonomy software stack optimized for inference.

Skills

ML fundamentals

Reinforcement learning

Frameworks: PyTorch

Frameworks: TensorFlow

Frameworks: Ray

Education

PhD or MS in CS/ML/Robotics

Tools

PyTorch

TensorFlow

Ray

Job description

Rainfall Ventures in San Francisco seeks a Machine Learning Engineer to design and advance a modular autonomous robot stack that fuses Vision-Language-Action models with purpose-built modules for grasping in unstructured environments.

You will implement action refinement and safety layers, architect interfaces for model swapping, and build robust data pipelines and high-speed inference software for production deployment.