Senior Embedded AI Developer

Senior Embedded AI Developer

BMW Group

Munich, Germany

What awaits you?

• You own the full lifecycle of machine learning models – from initial concept and experimentation through training and validation to reliable deployment on production-grade automotive systems.
• Furthermore, you optimise AI models for embedded platforms, driving performance analysis and benchmarking, as well as runtime, compute, and memory efficiency on next-generation edge hardware.
• You apply state-of-the-art compression techniques (quantisation, pruning, distillation, sparsity) to enable high-performance AI under strict real-time, energy, and hardware constraints.
• Additionally, you collaborate closely with software, machine learning, and semiconductor specialists to advance hardware/software co-design, ensuring functional correctness, safety, and deterministic real-time behaviour.
• You stay abreast of cutting-edge machine learning research and selectively translate promising developments into robust, scalable industrial applications on advanced automotive hardware.
• You design and operate scalable, MLOps-driven optimisation pipelines, enabling training, deployment, and inference across diverse embedded targets and accelerator architectures.

What should you bring along?

• University degree in electrical engineering, computer science, data science, artificial intelligence, mathematics, physics, or a comparable qualification.
• Extensive professional experience in the development, implementation, and operation of AI/ML models in an industrial or commercial environment.
• Strong expertise in embedded AI, particularly in neural network architectures, quantisation, hardware‑ and quantisation‑aware training, as well as deployment formats and cross‑platform portability.
• In‑depth expert knowledge of hardware/software co‑design for ML systems, including cross‑layer optimisation, architecture search, benchmarking, and the design of neural accelerators.
• Proven success in transferring ML models from research or prototype phases into productive, value‑creating applications on embedded targets.
• Comprehensive expertise in modern machine learning methods, including deep learning, multimodal models, end-to-end learning and transformer architectures, combined with a realistic understanding of their practical limitations.
• Very good English skills, both written and spoken, for working in an international environment; German skills are desirable and an advantage.