English summary for screening — check the original posting before applying.
This role focuses on developing foundational models for robot learning, starting from data collection and preparation, progressing through training and evaluation, and finally to model improvement using imitation and reinforcement learning. It's designed for individuals with a background in machine learning research or practical experience to contribute to real product development in the robotics domain.
Must-haves
- 2-3+ years of professional work experience
- Master's degree or higher in Machine Learning research OR practical experience in Machine Learning/data-related tasks at a company
- Practical Python experience
- Development experience using Git
- Ability to read technical documents and implement/verify them independently
Nice-to-haves
- Research or implementation experience in imitation learning, reinforcement learning, VLM/VLA, or Sim2Real
- Experience with LeRobot, Hugging Face Transformers, ROS/ROS2, Isaac Sim/Isaac Lab
- Development experience including real robots, sensors, and actuators
- Top rankings in machine learning competitions (e.g., Kaggle)
- Paper acceptance/presentation at conferences like CVPR, ICCV, NeurIPS, ICLR, IROS, CoRL, ICRA
Tech stack
PythonPyTorchLeRobotHugging FaceRSL RLNVIDIA Isaac LabNVIDIA Isaac SimDockerAWSGitSlackNotionClaude CodeCodex
Work style
Hybrid work is possible after the probation period. Location: Tokyo (Chiyoda-ku, Chuo-ku). Flexible working hours with core time 11:00-16:00.
Other notes
Estimated annual salary: 6 million - 10 million JPY (negotiable based on experience and ability). Probationary period: 3 months.