Principal Software Engineer, Perception Pretraining
Core
Developing sensor-fusion foundation models for the Waymo Driver to provide temporal-spatial understanding of the surrounding environment using millions of miles of driving data.
Role type
Principal Software Engineer, Perception Pretraining
Builds
Sensor-fusion foundation models for autonomous driving
Domain
Autonomous driving, computer vision, multi-modal foundation models
Deliverable
production ML models
Required skills
Large-scale vision/video/multi-modal foundation model development, ML-driven production systems, technical leadership, computer vision/LLM/vision language models, C++, Python, PyTorch/JAX
Preferred skills
Multi-modal LLM pretraining/fine-tuning, model distillation, compute optimization, end-to-end model development
Technologies
C++, Python, PyTorch, JAX
Responsibilities
Develop sensor-fusion foundation models, solve large-scale multi-stage training with limited labeling, update models to minimize time-to-impact, optimize end-to-end latency, drive collaborations between Perception, AI Foundation, Planning, Simulation, and Alphabet research teams
Seniority
Principal, technical leadership & strategy