Robotics Engineer

Location: Munich, Germany (Hybrid)Department: Advanced R&DExperience Level: Senior / Lead (5+ years)Key Tech Stack: PyTorch, ROS 2, C++, Python, Isaac Sim / MuJoCoLanguage: German (Fluent)
The OpportunityWe are collaborating with a stealth-mode innovator backed by global technology leaders, currently building an elite core team in Munich. This organization operates at the intersection of deep learning and physical actuation, developing next-generation foundation models specifically optimized for complex, multi-modal physical interactions.
By decoupling from rigid legacy architectures, the team focuses purely on generalizable, adaptive intelligence deployed onto state-of-the-art kinematically diverse robotic systems. This role offers an unprecedented opportunity to influence the architectural landscape of physical AI from the ground up in continental Europe's primary deep-tech hub.
Role & ResponsibilitiesAs a Senior Specialist in Embodied AI, you will lead the bridge between advanced machine learning constructs and real-world physical deployment. Your core focus will be expanding behavioural generalization across multi-modal architectures.

• Foundation Model Engineering: Design, train, and fine-tune Vision-Language-Action (VLA) and deep transformer-based models capable of zero-shot generalization to novel tasks and physical environments.
• Sim-to-Real Transferability: Optimize physical and visual simulation frameworks (e.g., Isaac Sim, MuJoCo) to dramatically minimize the sim-to-real gap, deploying learned physical behaviours seamlessly onto hardware.
• Robot Learning Pipelines: Develop scalable infrastructure for imitation learning, behaviour cloning, and reinforcement learning (RL) using extensive, diverse real-world datasets.
• Algorithm Deployment: Optimize large-scale deep learning models for real-time inference on edge compute structures, ensuring latency constraint boundaries are strictly met.
• Cross-Functional Collaboration: Work intimately with deep-learning engineers, mechatronics designers, and perception teams to co-design hardware-software requirements for agility, manipulation, and perception.

Profile & QualificationsWe are looking for an analytical, execution-oriented engineer who thrives in ambiguity and values tangible, physical deployment over theoretical benchmarks alone.

• Education: Master’s degree or PhD in Computer Science, Robotics, Machine Learning, Cognitive Systems, or a closely related quantitative field.
• Core Technical Expertise: Strong foundation in deep neural networks, transformers, and continuous control algorithms. Proven experience with robotic kinematics, manipulation, or dynamic locomotion is highly desirable.
• Programming Excellence: Exceptional production-level coding skills in Python and modern C++ (17/20), along with standard frameworks like PyTorch or JAX.
• Robotics Ecosystems: Deep structural familiarity with ROS 2 (Robot Operating System) and standard simulation tools.
• Mindset: A passionate builder who wants to see their algorithms manifest in physical world manipulation and movement, comfortable working in rapid sprint cycles.

What We Offer

• Highly competitive compensation package including equity/virtual share options tailored to early-stage growth.
• Direct access to cutting-edge, proprietary robotic hardware and massive localized compute clusters.
• A vibrant office location in the heart of Munich with flexible hybrid arrangements.
• Relocation assistance to Germany for international candidates, alongside extensive visa support.
• An intellectually rigorous environment consisting of international experts from top academic institutions and elite global tech labs.