Muso Action
/
Careers

Full-time

AI Robotics Engineer

About the role

This role builds the “brain” of Muso-roid, our general-purpose robot worker. You own the full learning stack of physical AI — from imitation- and reinforcement-learning models to data collection and simulation environments.

Responsibilities

  • Implement and evaluate imitation-learning (VLA, ACT, etc.) and reinforcement-learning (PPO, etc.) models
  • Robotics R&D: physics-simulator environments, CV preprocessing for model training, PID tuning, and more
  • Apply human egocentric data to robot learning; paper surveys and prototyping around WAM techniques

Why this role

  • Put state-of-the-art VLA and imitation learning to work on real robots, not just in papers
  • Design the entire learning pipeline — models through data collection — with real ownership
  • Shape technology choices and engineering culture as a founding-phase engineer

Required

  • Hands-on implementation/evaluation of imitation or reinforcement learning on real robots
  • Experience building physics-simulator environments or data-collection pipelines for model training
  • Development experience in Python and C++
  • Development experience on Linux
  • Experience developing with AI coding tools

Nice to have

  • Development/evaluation experience with LeRobot / Isaac Lab
  • Manipulation-model development using human egocentric data
  • WAM development experience
  • Hands-on training/inference experience with PyTorch or similar

Who we're looking for

  • You insist on building robots that are truly used and relied on in the field — not built for demos
  • You enjoy the absence of structure, and can design and establish things from zero to one
  • You welcome autonomy and speed in a small team, and drive yourself forward

Employment & salary

  • Employment: full-time
  • Salary: ¥8M–¥20M (depending on experience and skills)
  • Discretionary labor system for specialized work

Location

Daini TOC Bldg. 6F, 7-21-11 Nishi-Gotanda, Shinagawa-ku, Tokyo

Benefits

  • Full social insurance
  • Two days off per week (Sat/Sun) and national holidays
  • 20 days of paid leave, granted from the first year

Apply for this position

Apply using the form below. We will review your documents and get back to you.

Screening questions

Please choose the option that best fits you for each requirement of this position.

Q1. Experience implementing and evaluating imitation or reinforcement learning on real robots?
Q2. Experience building physics-simulator environments or data-collection pipelines?
Q3. What is your professional development experience with Python and C++?

PDF / Word, up to 3MB each

Submitted documents are used only for recruiting purposes.