Staff ML Lead: Perception Pre-Training & Distillation

Waymo; Mountain View (CA); On-site; USD 251,000 - 310,000; Full time

Benefits offered by this job

Annual bonus

Equity incentive

Benefits program

Job summary

Waymo is seeking a senior machine learning/computer vision researcher in Mountain View to advance the Waymo Driver. You will train multi-billion parameter models, develop self-supervised learning methods, and deliver pre-trained models for real-time deployment.

You will also create distillation recipes, evaluate representations, and curate diverse training datasets while contributing to technical roadmaps. The role requires strong Python, ML framework experience (PyTorch/JAX), and a track record

Qualifications

• MS or PhD in Computer Science or a closely related field, or equivalent deep learning experience.
• 5+ years of experience in Machine Learning and/or Computer Vision.
• Fluent in Python.
• In-depth experience with ML frameworks like PyTorch or JAX.
• Publications at top-tier conferences (preferred).
• Experience with C++ (preferred).
• Experience with real-time system design (preferred).

Responsibilities

• Train multi-billion parameter spatio-temporal foundation models on multi-billion frame datasets.
• Deliver pre-trained models for production usage, ideally for real-time systems.
• Develop self-supervised learning, including image/video encoder, 3D point cloud encoder, and spatio-temporal tasks.
• Develop multi-teacher distillation recipes.
• Develop methods to understand and evaluate latent representations from pre-training; collaborate with post-training team.
• Develop methods to curate and effectively use diverse datasets in training.
• Be a key contributor to technical roadmaps.

Skills

Python

Machine Learning

Computer Vision

Deep Learning

Education

MS/PhD in Computer Science or related

Tools

PyTorch

JAX

C++

Real-time systems

Job description

Waymo is seeking a senior machine learning/computer vision researcher in Mountain View to advance the Waymo Driver. You will train multi-billion parameter models, develop self-supervised learning methods, and deliver pre-trained models for real-time deployment.

You will also create distillation recipes, evaluate representations, and curate diverse training datasets while contributing to technical roadmaps. The role requires strong Python, ML framework experience (PyTorch/JAX), and a track record