Robotics Engineer, Foundation Model (Open Position)

29 days ago

Tokyo, JapanMid Level / Senior / Staff+

Responsibilities

Design, train, and fine-tune large-scale multimodal, transformer-based, and Vision-Language-Action foundation models for robotic perception, reasoning, and control.
Build and maintain data and training pipelines covering teleoperation data collection, preprocessing, annotation, and distributed training.
Integrate models with real robot systems and evaluate their performance in production environments alongside robotics, controls, and hardware engineers.
Design evaluation protocols for model performance, safety, and reliability, and iterate based on real-world deployment feedback.
Optimize models for efficient inference and deployment on embedded and edge hardware.
Track foundation model, LLM, multimodal, and generative AI developments and assess their applicability to products.

Background in machine learning, computer science, robotics, or a related field.
Professional experience in machine learning or deep learning engineering, or equivalent research or graduate experience.
Hands-on experience training, fine-tuning, or serving large-scale models such as LLMs, vision-language models, diffusion models, or other multimodal foundation models.
Strong software engineering skills in Python and experience with a deep learning framework, preferably PyTorch.
Familiarity with large-scale or distributed training and modern ML infrastructure or MLOps practices.
Strong problem-solving skills and ability to work in cross-functional teams.
Preferred experience includes robotics with ROS or ROS2, reinforcement learning, embodied AI, Vision-Language-Action or other multimodal foundation models, NVIDIA Jetson edge deployment, computer vision, NLP, generative modeling research, control theory, teleoperation systems, actuator or hardware integration, publications, open-source contributions, or an applied ML/AI portfolio.
Professional proficiency in English is required; Japanese language skills are a plus.

Opportunity to develop foundation models that control real robots in stores and warehouses.
Collaborate with multidisciplinary teams to bring physical AI from prototype to real-world deployment.
The role is open across wide career stages, with scope and responsibility aligned to experience and expertise.

51-200 employees