Key Notes
Release ownership spans Wayve’s multi-stage ML training lifecycle, with responsibility for standards at each gate before models progress. The role calls for MLOps and model-lifecycle experience, deep ML-training knowledge, ML code infrastructure and CI/CD with GitHub Actions. Relevant technologies include PyTorch, TensorRT, quantisation and deployment. The London role is full time under Wayve’s hybrid policy.
What You'll Work On
- Set and enforce release-gate standards across Wayve’s multi-phase machine-learning training and model-delivery lifecycle.
- Review model changes, metric changes and evaluation results against Wayve’s quality and safety standards before release.
- Identify ML delivery-pipeline bottlenecks and work with AI Platform teams on checks and automation that surface issues earlier.
- Adapt CI/CD workflows with platform teams to streamline model delivery without weakening release standards.
- Improve model-evaluation methodology with evaluation teams and introduce relevant MLOps practices and tooling into the workflow.
Why This Role Matters
- Wayve’s model baseline is protected by release gates across successive training phases, so validation decisions determine whether model and metric changes are allowed to progress through the delivery lifecycle.
- On-road product performance depends on released models meeting Wayve’s stated quality and safety standards, making evaluation reliability and release-content review consequential before models ship.
What They Are Looking For
- Experience: Experience introducing operational processes that build engineering excellence is essential.
- Technical: Strong MLOps, model-registry and ML-lifecycle experience plus deep knowledge of ML training and ML code infrastructure are essential.
- Technical: Relevant technical skills include PyTorch, TensorRT, quantisation and model deployment.
- Technical: Strong CI/CD and GitHub Actions experience is essential.
- Preferred: Grafana monitoring and production-observability experience is desirable.