Machine Learning Engineer, Compiler

b Overview /b p In this role, you own the ML compilation pipeline from model checkpoint to deployable bundle for embedded targets. You will design compiler passes and build scalable infrastructure that spans NVIDIA TensorRT and Qualcomm QNN, ensuring accuracy and latency targets are met. You collaborate with model and training teams to validate changes and push the compiler stack across architectures. Your work directly influences what runs on Wayve’s embedded driving hardware, enabling efficient, reproducible deployment. This is a hands-on position with strategic impact on the compiler ecosystem and product readiness. /p b Responsibilities /b ul li Own the end-to-end ML compilation pipeline for NVIDIA TensorRT and Qualcomm QNN targets /li li Design and implement compiler passes with accuracy and latency gates /li li Build scalable compilation infrastructure across platforms and SoCs /li li Partner with model and training teams on compilability; build regression and benchmarking /li li Set technical direction and mentor on compiler design /li /ul b Key requirements /b ul li Built or owned significant parts of an ML compilation or graph-lowering pipeline /li li Deep experience with quantisation in compilation (precision typing, PTQ integration, debugging accuracy loss) /li li Strong Python; comfortable developing and testing compiler infrastructure in production codebases /li li Proficiency in MLIR, ONNX, TensorRT, Qualcomm QNN, or PyTorch graph capture/export /li li Experience with multi-target compilation or graph partitioning across hardware backends /li li Ability to reason about correctness and performance trade-offs at each compiler stage /li /ul ul li Clear communicator to align cross-functional teams /li li Mentoring on compiler design /li li Strategic thinking with ability to set direction /li li MLIR /li li ONNX /li li TensorRT /li /ul