The Labeling Platform Team creates data solutions to power groundbreaking research and development during all stages of model training: pretraining, supervised fine-tuning, reinforcement learning, etc. The labeled data that the team produces is used to directly enhance and evaluate the Waymo Driver as well as the vast variety of models that power other parts of the business.

You Will:

• Build a Data Flywheel: Design and implement systems that continuously improve the Waymo Driver. This involves creating a feedback loop where labeled data not only trains models but also informs the next generation of data collection and labeling, leading to iterative improvements in quality and efficiency.
• Collaborate with ML practitioners on models for perception, behavior prediction and planner  to understand their models and help them define data requirement,
• Develop automation to both produce the labels and/or find problematic label issues within the already produced labeled datasets.
• Build large scale distributed systems that index and serve the large amount of labeled data to model training pipelines.

You Have:

• Bachelor's degree in Computer Science, Engineering, or related field, and 2+ years equivalent experience
• Experience in producing and managing large scale ML training datasets.
• ML Model development experience
• Experience in programming C++ or Python.

We Prefer:

• Experience in developing human + ML interaction schemes to boost productivity and quality of label generation.
• Experiencing programming using Typescript.
• Hands-on experience with LLM/GenAI-based products.
• Experience using open source computer vision models such as YOLO and SAM.

This role reports to an Engineering Manager

#LI-Hybrid