Key Notes
Eight-plus years in ML engineering, including production computer vision, and at least two years leading or managing senior engineers are essential. This London hybrid role manages four senior ML engineers and owns offline scene-understanding models for counterfactual evaluation, rare-event mining and AV2.0 assessment. Essential depth covers foundation models, multimodal architectures, large-scale training, Python and ML frameworks, especially PyTorch.
What You'll Work On
- Lead four Senior Machine Learning Engineers and hire complementary ML, software and data-science specialists as the Measurement team expands.
- Set the technical architecture for adapting on-vehicle and Wayve Foundation Models into scalable offline scene-understanding production systems.
- Own sprint, quarterly and annual roadmaps that translate ambiguous evaluation goals into concrete technical programmes and investments.
- Establish production engineering standards for customer-facing ML deliverables, rig-agnostic architectures, monitored systems and clean technical interfaces.
- Align UK and US teams across AV Core, Foundation Models, Evaluation, Simulation and Model Development Platform roadmaps.
Why This Role Matters
- Wayve’s model-development decisions and customer deliverables depend on offline scene-understanding models that assess counterfactual outcomes, coverage and rare events after on-road runs and simulation.
- The Measurement team’s production models connect driving-model iterations to reliable evaluation feedback, influencing how quickly engineers receive actionable results for the next development cycle.
What They Are Looking For
- Experience: Eight-plus years in ML engineering, including computer vision with camera and/or lidar data and delivery of monitored, customer-facing production systems, is essential.
- Experience: Two-plus years line-managing or technically leading senior engineers, including hiring, development and retention, is essential.
- Technical: Transformer-based and multimodal architectures, foundation models, large-scale training and staff-level design review are essential.
- Technical: Python and ML frameworks, especially PyTorch, with production-grade engineering judgement are essential.
- Stakeholder: Cross-functional roadmap alignment across teams, geographies, engineers and senior leadership is essential.
- Preferred: Autonomy perception, 3D scene understanding, offline models, counterfactual evaluation, distributed teams or a relevant postgraduate degree is desirable.