12 hours ago
Responsibilities
- Own the ML compilation pipeline from model checkpoint to deployable bundle for NVIDIA TensorRT and Qualcomm QNN targets.
- Design and implement compiler passes covering graph lowering, precision assignment, legalisation, partitioning, and related stages.
- Build accuracy and latency gates, regression testing, and benchmarking to validate compiler changes across releases.
- Develop scalable compilation infrastructure for multiple model architectures, platforms, and SoCs.
- Partner with model and training teams on compilability and resolve compiler trade-offs affecting correctness and performance.
- Set technical direction and mentor others on compiler design.
Requirements
- Experience building or owning significant ML compilation or graph-lowering pipelines.
- Deep experience with quantisation in compilation, including precision typing, post-training quantisation integration, and debugging accuracy loss from compiler transforms.
- Strong Python proficiency and experience building and testing production compiler infrastructure.
- Proficiency with at least one of MLIR, ONNX, NVIDIA TensorRT, Qualcomm QNN, or PyTorch graph capture/export.
- Experience with multi-target compilation or graph partitioning across hardware backends.
- Ability to reason about correctness, latency, accuracy, and vendor-backend constraints across compiler stages.
- C++ experience is a plus, along with clear communication and cross-functional collaboration skills.
Benefits
- Inclusive interview accommodations are available upon request.
- The role contributes to autonomous driving products running on real vehicle hardware.
- The posting describes a greenfield, high-leverage Staff-level opportunity with technical direction and mentoring scope.
Categories
About Wayve
Wayve builds end-to-end autonomous driving software—the vehicle-agnostic Wayve AI Driver—that runs on onboard compute and native sensors, licensed to automakers and fleet operators. Its platform spans ADAS and higher autonomy (L2+/L3 to robotaxi) and is designed to generalize across vehicle types and geographies. Founded in 2017 and headquartered in London, it tests its models across Europe, North America, and Japan, with a U.S. base in Sunnyvale, CA.