Main Responsibilities
- Build and ship production AI models and training pipelines for robotics and autonomous systems.
- Design, train, and scale large multimodal foundation models, including VLMs, world models, vision-language-action models (VLAs), 3D scene reconstruction (e.g., 3D Gaussian Splatting), perception systems, and data curation frameworks.
- Develop and contribute to open-source codebases, tooling, and reference implementations to accelerate adoption and collaboration.
- Advance the state of the art through publications and open model/code releases.
Collaboration
- AI Engineering Teams: Advance state-of-the-art perception and world models while building reusable training and evaluation pipelines.
- Product and Software Engineering Teams: Co-design and integrate AI workloads into products; support CI/CD validation and provide actionable feedback on architecture and interfaces.
- Internal and External Partners: Align technical roadmaps with academic, industrial, and internal stakeholders; translate objectives into practical plans and milestones.
- Cross-Site Teams: Collaborate effectively with product and research teams across Europe and globally, ensuring clear ownership, strong feedback loops, and predictable delivery.
Goals for the First 6 Months
- Become proficient with the codebase, infrastructure, and end-to-end pipelines being developed for Physical AI applications.
- Assess emerging developments in the field and translate insights into engineering priorities and a model roadmap.
- Train and release an end-to-end model for a perception, world-model, or autonomous-systems use case, including training, evaluation, and deployment.
- Lead client-facing implementation efforts to identify product needs and convert them into actionable technical deliverables.
Ideal Candidate Profile
Skills and Qualifications
- Master's degree, PhD, or equivalent experience in Machine Learning, Robotics, or a related field, including 3+ years of relevant industry experience.
- Experience building robotics, perception, or autonomous systems pipelines.
- Strong foundation in deep learning for perception and embodied decision-making, including transformers, diffusion models, and world models.
- Hands-on experience with vision and multimodal foundation models (e.g., ViT, CLIP, DINO, LLaVA) and VLAs (e.g., OpenVLA, Pi-0.5).
- Experience using simulation environments and RL/IL techniques to train and evaluate embodied agents (e.g., Isaac Lab, MuJoCo, Genesis, LeRobot).
- Proficiency in Python and familiarity with C++ in production environments.
- Strong experience with PyTorch; experience with JAX is a plus.
- Strong engineering skills, including rapid prototyping, debugging, profiling, optimization, AI-assisted development tools (e.g., Claude Code, Cursor), and delivering maintainable production code.
Preferred Qualifications
- Experience building and operating large-scale machine learning systems, including training infrastructure and distributed computing environments.
- Publication record in leading conferences such as CVPR, ICCV, ECCV, NeurIPS, ICRA, or IROS.
- Experience profiling and optimizing GPU workloads using ROCm and/or CUDA.
- Experience developing, maintaining, and supporting open-source software projects, including releases, documentation, and CI pipelines.
- Experience with cloud platforms (AWS, GCP, Azure) and cluster orchestration technologies such as Slurm, Kubernetes, or Yarn.
Location
Sweden, Finland, Germany or UK
#LI-MH3
#LI-HYBRID