VLA Robotics Engineer (m/f/d)

Agile Robots SE; München; Vor Ort; EUR 70.000 - 95.000; Vollzeit

Zusammenfassung

Agile Robots SE seeks a Robotics Engineer to apply modern AI models to physical robotic systems. You will develop, test, and deploy Vision-Language-Action driven capabilities on real production lines, bridging AI research and practical execution.

You will work with ROS/ROS2, Python/C++, and ML frameworks on edge hardware, supporting data collection and model iteration. Occasional on-site work in Munich with travel to China may be required.

Qualifikationen

• 2 to 4 years of professional or practical experience in robotics or ML.
• Experience applying AI models to physical systems is required.

Aufgaben

• Implement, fine-tune, and evaluate Vision-Language-Action models and multimodal policies for robotic manipulation tasks.
• Deploy and validate robotic control pipelines on physical hardware in production environments.
• Integrate multimodal model outputs into existing robot execution and control frameworks.
• Assist in data collection campaigns and set up data pipelines for model training.
• Test model performance on edge hardware ensuring system reliability.
• Write clean, maintainable code and documentation to support CI/CD.
• Travel to deployment sites, mainly China, up to 30%-40% of time, to work on-site.

Kenntnisse

Vision-Language-Action models

Robotics fundamentals

Problem solving

English communication

Writing readable/modular code

Tools

Python

C++

ROS/ROS2

Linux

Docker

Git

ML frameworks (e.g., PyTorch)

Jobbeschreibung

The AI team at Agile Robots is seeking a motivated Robotics Engineer with experience applying modern AI models to physical systems. You should have 2 to 4 years of professional or practical experience in robotics and machine learning, with a strong interest or hands-on experience in Vision-Language-Action (VLA) models. In this role, you will help bridge the gap between AI research and practical execution by implementing, testing, and deploying VLA-driven capabilities directly onto real-world production lines. Prior experience running learning-based policies on real hardware in industrial setups is a significant plus.

Your Responsibilities

• Implement, fine-tune, and evaluate Vision-Language-Action (VLA) models and multimodal policies for robotic manipulation tasks
• Deploy and validate robotic control pipelines on physical hardware in production-line and factory environments
• Integrate multimodal model outputs (vision, force/torque, language inputs) into existing robot execution and control frameworks
• Assist in data collection campaigns, recording robot demonstrations, and setting up data pipelines for model training and iteration
• Test model performance on edge hardware, ensuring consistent performance of the system
• Write clean, maintainable code, test scripts, and technical documentation to support continuous integration and deployment
• Travel to our deployment sites, primarily China (up to 30%-40% of the time depending on project stage), to work alongside on-site engineers, deploy solutions on production floors,
• and gather operational feedback
• 2 to 4 years of relevant professional or applied experience in robotics software, computer vision, or machine learning for robot control
• Solid understanding of Vision-Language-Action models, multimodal transformers, or learning-based manipulation (imitation/reinforcement learning)
• Proficient in writing modular, readable, and well-tested code in Python and/or C++.
• Hands-on experience with ROS/ROS2, Linux environments, Docker, Git, and ML frameworks
• like PyTorch
• Solid understanding of robotics fundamentals, including kinematics, dynamics, control, and
• motion planning
• Comfortable working directly with physical robots, sensor integration (cameras, force
• sensors), and basic control loops
• Availability to work primarily on-site at our headquarters in Munich
• Strong problem-solving skills and good written and verbal communication in English

Beneficial Skills

• Hands-on experience testing or deploying learned policies in actual production lines or industrial factory setups
• Experience with robot demonstration processing, data logging, and data preparation for robot learning pipelines
• Good technical writing skills for clear setup guides and deployment documentation
• Good verbal and written communication in Chinese (Mandarin) is a strong plus for smooth on-site collaboration at our Chinese deployment facilities

What we offer

• A dynamic high-tech startup, combined with financial soundness and world-class investors
• Challenging tasks and projects
• Great working environment in an interdisciplinary, international team
• Room for creativity and personal development