Latest AI JobsBrowse Jobs
rohlik logo

Robot Learning Engineer

rohlikMünchen, GermanyIT
Apply on company site
Salary
Not listed
Job type
Full-time
Work mode
Remote
Experience
Mid · 3+ yrs
Posted
22 Sept 2026
Status
Open · checked 7d ago

About this role

Robot Learning Engineer at rohlik is an opportunity based in München, Germany, focused on building physical AI systems to automate fulfilment centres. This position suits an experienced machine learning professional who wants to take ownership of capture specifications, data evaluation, and initial robot policy training.

Role Overview

Robot Learning Engineer is an engineering position within a small, newly created team at rohlik in München, Germany. The successful candidate will act as the machine learning voice in the room, working alongside an operations lead and a data engineer. The core mission involves recording human handling of groceries, transforming those recordings into a robust training corpus, and post-training open robot-learning models.

Key Responsibilities

Define the capture specification covering cameras, mounts, calibration, and disqualification criteria for episodes.
Own the measurement and closing of the gap for ground truth, hand pose, and warehouse reality factors like work gloves and occlusion.
Build the evaluation harness to determine dataset trainability and maintain the quality bar as data volume grows.
Post-train open robot-learning models using collected data and execute initial task evaluations.
Review published research, reproduce relevant claims, and guide hardware capture spending decisions.

Requirements

Three or more years of relevant full-time robotics experience.
Proficiency in PyTorch and computer vision involving real 3D geometry, including camera calibration, pose estimation, and SLAM fundamentals.
Demonstrated ability to read and reproduce published academic papers.
Familiarity with the current robot-learning stack and policy classes to anticipate training demands.
Experience training models on self-collected data and understanding associated failure modes.
Comfort working as the sole machine learning voice and directing AI agents for plumbing and scripting tasks.

Nice to Have

Hand-pose estimation or egocentric video in production.
Multi-camera rigs and time synchronization experience.
Background in post-training robot-learning models.
Familiarity with ROS 2.

Skills and Tools

PyTorch
ROS 2
Computer Vision
3D Geometry
Camera Calibration
Pose Estimation
SLAM
Claude Code
Devin

About the Company

rohlik operates as a leading Central European e-grocer serving over a million customers across multiple regional platforms including Rohlik.cz, Knuspr.de, Kifli.hu, Gurkerl.at, and Sezamo.ro. The company delivers over 17,000 items within short time windows and is actively building an autonomous grocery fulfilment network utilizing AI assistants, vision models, and physical robotics.

How to Apply

Interested candidates should visit the official rohlik careers page to submit their application.

Skills and tech stack

PyTorchROS 2Claude CodeDevinComputer VisionMachine Learning

Questions candidates ask

Where is the Robot Learning Engineer position located?

The role is located in München, Germany.

How many years of experience are required for this role?

The posting requires 3+ years of relevant full-time robotics experience.

Is a publication record required for this position?

No publication record is required, as this is not a research-scientist role.

What specific technical stack and tools are mentioned?

The technical stack includes PyTorch, ROS 2, computer vision tools for 3D geometry, camera calibration, pose estimation, SLAM, along with AI agents like Claude Code and Devin.

What team structure will the engineer work within?

The engineer will be part of a small, newly created team consisting of an operations lead, a data engineer, and the machine learning engineer.

Ready to apply?

The application is on the official rohlik careers page.

Apply on company site