Robotics ML Research Engineer

Relari, inc.; San Francisco (CA); On-site; USD 125,000 - 185,000; Full time

Job summary

Relari, inc. in San Francisco develops machine learning systems for dexterous robotic manipulation. We seek engineers who turn research ideas into working systems, spanning data processing, training, evaluation, and real-world deployment.

You will design and optimize training pipelines, build scalable infrastructure, and collaborate with robotics engineers to deploy models on physical robots using feedback from real-world use.

Qualifications

• Strong Python programming skills and experience with ML frameworks.
• Deep learning, model training, optimization, and evaluation fundamentals.
• Experience building ML systems beyond APIs and pretrained models.

Responsibilities

• Develop and improve training pipelines for models using human and sensor data.
• Train and adapt models for manipulation and imitation learning.
• Explore representations from vision, motion, and signals.
• Experiment with imitation learning, RL, and foundation models for robotics.
• Design experiments to understand model behavior and improvements.
• Build scalable data processing, training, and evaluation infrastructure.
• Collaborate with robotics engineers to deploy models on physical robots.

Skills

Python

PyTorch

JAX

Deep learning

Experiment design

ML infrastructure

Robotics

Computer vision

Job description

Relari, inc. in San Francisco develops machine learning systems for dexterous robotic manipulation. We seek engineers who turn research ideas into working systems, spanning data processing, training, evaluation, and real-world deployment.

You will design and optimize training pipelines, build scalable infrastructure, and collaborate with robotics engineers to deploy models on physical robots using feedback from real-world use.