Machine Learning Engineer (Vision Foundation Models)

Location: Hybrid in Foster City, CA; Boston, MA; Seattle, WA; or San Diego, CA, United States

Company

Zoox develops purpose-built robotaxis and the autonomy software that enables them to understand and safely navigate the world.

What you will do

• Design, train, and evaluate vision models for in-cabin passenger, object, and behavior understanding, including detection, pose, and action recognition.
• Own a model or feature area end to end, from data curation, labeling guidelines, and training through evaluation and deployment.
• Set quality standards through rigorous metric design and error analysis for models supporting live operations.
• Deploy models under real-time and compute constraints within the onboard perception system.
• Partner with product, operations, safety, and vehicle software teams to define model specifications, metrics, and validation plans.

Requirements

• MS or PhD in Computer Science, Machine Learning, Computer Vision, or a related field.
• 6+ years of industry experience with an MS, or 3+ years with a PhD.
• Deep expertise in computer vision and deep learning for image and video understanding, including object detection, pose and action recognition, and fine-grained classification.
• Hands-on experience with convolutional and vision-transformer detectors and self-supervised or foundation-model backbones such as DINOv3.
• Strong proficiency in Python and PyTorch for model design, training, and evaluation.
• Experience with vision-language models or foundation-model backbones, and with auto-labeling or synthetic data for long-tail and rare classes.
• Track record of owning the full ML lifecycle on real-world data, including dataset curation, labeling guidelines, training, precision/recall/false-positive-rate metric design at an operating point, and error analysis.

Nice to have

• Experience with in-cabin or occupant monitoring systems, including multi-camera or fisheye imagery and low-light conditions.
• Experience deploying deep learning models to real-time, resource-constrained robotics, automotive, or edge systems using TensorRT, ONNX, or similar tools.
• Proficiency in C++ for model integration.

Culture & Benefits

• Full-time hybrid work arrangement.
• Models directly support rider safety and the experience of operating a robotaxi service.
• Zoox welcomes applicants who do not match every listed expectation and values diverse backgrounds, experiences, and skills.