Machine Learning Engineer, ADAS at Wayve
Germany
Posted
on-site · full-time · Visa sponsorship
Responsibilities:
- Train, debug, and improve computer vision and 3D perception models for ADAS (detection, classification, instance segmentation).
- Own the full ML lifecycle: data collection, labelling (auto-/pseudo-labelling), training, evaluation and iteration.
- Build and maintain scalable data pipelines and tooling to generate large, high-quality datasets (tracking, multi-frame reconstruction).
- Develop offline reconstruction and tracking systems to propagate labels through time and improve dataset coverage.
- Work on both online (latency/compute-constrained in-car) and offline models, optimizing trade-offs for deployment.
- Collaborate cross-functionally with engineers, data scientists and product to prioritise model improvements and integrate solutions.
- Produce clear evaluation metrics and diagnostic tooling to drive iteration and measurable impact.
Requirements:
- Proven applied ML engineering experience: built and shipped CV-focused deep learning systems (not research-only).
- Practical experience with 3D perception concepts: LiDAR, multi-view geometry, tracking, or 3D reconstruction.
- Strong software engineering skills in Python and ML frameworks (PyTorch/TF), and experience building data pipelines/automation.
- Familiarity with auto-labelling / pseudo-labelling techniques and large-scale dataset generation.
- Experience balancing model accuracy with runtime constraints for deployment on vehicles is a plus.
- Good communication, ownership of end-to-end features, and ability to work in fast-paced, ambiguous environments.