Robot Learning Engineer

The role

You will create a training corpus for manipulation policies by defining capture specifications and ground truth measurements. You will build an evaluation harness and perform the first training runs on robot-learning models using collected data.

What you will do

• Write the capture spec for cameras, mounts, and calibration.
• Own how hand pose is measured to close the gap between human video and robot policies.
• Build the evaluation harness to decide which hours of footage are trainable.
• Post-train open robot-learning models on collected data.
• Run the first task evaluations.
• Reproduce claims from published work to steer capture before spending money.

Required qualifications

• 3+ years of relevant full-time robotics experience.
• PyTorch.
• Computer vision with real 3D geometry.
• Camera calibration.
• Pose estimation.
• SLAM fundamentals.
• Ability to read a paper and reproduce it.
• Familiarity with the current robot-learning stack and policy classes.
• Experience training on data you collected yourself.

Nice to have

• Hand-pose estimation or egocentric video in production.
• Multi-camera rigs and time sync.
• Experience post-training robot-learning models.
• ROS 2.

About the company

Rohlik is the leading Central European e-grocer. They are building an autonomous grocer using AI assistants, agents, and robots in fulfillment centres.