Software Engineer, ML Infrastructure, Optimization
Core
Building and optimizing core infrastructure for machine learning model life cycles in self-driving vehicles, focusing on model compilers, quantization, and low-precision inference.
Role type
ML Infrastructure Engineer (Optimization & Compilers)
Builds
Optimized ML models and a world-class model compiler framework (FTL) for Nuro's autonomous robot fleet.
Domain
Autonomous driving / Robotics / Physical AI
Deliverable
production ML models | infrastructure
Required skills
ML optimization techniques (quantization, pruning), ML compilers, GPU runtime profiling, Python, C++, CUDA, deep learning frameworks (PyTorch, Jax, Tensorflow, Keras)
Preferred skills
Experience with large language models, end-to-end learned ML solutions
Technologies
FTL, PyTorch, Jax, Tensorflow, Keras, CUDA
Responsibilities
Optimize autonomy stack with quantization and low-precision inference; Develop and maintain the FTL model compiler framework; Collaborate with domain experts to implement end-to-end learned ML solutions; Write robust software to validate vehicle navigation safety.
Seniority
Mid-level, hands-on IC