🕒 7 days ago
Train and ship computer vision and 3D perception models for ADAS, owning work across data, evaluation, and iteration to improve real-world driving performance under production constraints.
Senior / Staff Machine Learning Engineer - Behavior Models for Road Users
📅 Dec 17, 2025
Machine LearningDeep LearningImitation LearningReinforcement Learning
📜 Description
- Develop deep learning models using imitation and reinforcement learning to generate driving plans for human-like agents.
- Build generative behavior models, including diffusion and autoregressive models, conditioned on scenario intent for targeted stress-testing.
- Develop techniques to assess driving plans for safety, progress, comfort, and realism.
- Create metrics and tools to analyze errors and measure system improvements.
- Collaborate with Perception, Planning, Simulation, and Validation engineers on autonomous driving.
🛠️ Requirements
- PhD degree in computer science or related field and 4+ years of relevant professional experience or master's degree and 7+ years of relevant professional experience
- Experience in one of the following: Planning, Prediction, Reinforcement Learning, Imitation Learning, generative modeling (diffusion, autoregressive models)
- Experience with training and deploying transformer-based model architectures
- Experience with production Machine Learning pipelines: dataset creation, training frameworks, metrics pipelines
- Fluency in Python ML frameworks and a basic understanding of C++
- Top tier publications (NeurIPS, ICML, CVPR)
- Experience with JAX
✨ Benefits
- Sick leave
- Vacation
- Bereavement leave
- Unpaid time off.
- Zoox Stock Appreciation Rights.
- Amazon Restricted Stock Units (RSUs).
- Health insurance.
- Long-term care insurance.
- Long-term and short-term disability insurance.
- Life insurance.
Full job description
This role in the Offline Driving Intelligence team is responsible for developing/learning behavior models for road users such as cars, bicycles, and pedestrians. These agents populate Zoox's simulations and must be indistinguishable from real road users, yet fully controllable: promptable into the rare, adversarial, safety-critical behaviors we need to test against. This means the team’s models directly impact how fast Zoox can train, validate and ship its driving stack. Our team collaborates closely with Planner, Simulation and Validation teams to develop and validate our driving performance. As an ML Agents Machine Learning Engineer, you will work on the bleeding edge of the industry, developing novel machine learning pipelines and models to predict the behavior of other agents in the world and planning the best course of action for the ego vehicle.
There are three major components to compensation for this position: salary, Amazon Restricted Stock Units (RSUs), and Zoox Stock Appreciation Rights. A sign-on bonus may be offered as part of the compensation package. The listed range applies only to the base salary. Compensation will vary based on geographic location and level. Leveling, as well as positioning within a level, is determined by a range of factors, including, but not limited to, a candidate's relevant years of experience, domain knowledge, and interview performance. The salary range listed in this posting is representative of the range of levels Zoox is considering for this position.
Zoox also offers a comprehensive package of benefits, including paid time off (e.g. sick leave, vacation, bereavement), unpaid time off, Zoox Stock Appreciation Rights, Amazon RSUs, health insurance, long-term care insurance, long-term and short-term disability insurance, and life insurance.
In this role, you will...
Develop new deep learning models that use imitation learning and reinforcement learning to generate driving plans for human-like driving agents.
Work on novel techniques to estimate the quality of those driving plans along the dimensions of safety, progress, comfort and realism.
Build generative behavior models (e.g. autoregressive, diffusion) that are conditionable on scenario intent — "cut off the ego vehicle," "jaywalk here" — for targeted stress-testing.
Leverage our compute, infrastructure and large corpus of data to push boundaries of the field.
Develop metrics and tools to analyze errors and understand improvements of our systems.
Collaborate with engineers on Perception, Planning, Simulation, and Validation to solve the overall Autonomous Driving problem.
Qualifications
PhD degree in computer science or related field and 4+ years of relevant professional experience or master's degree and 7+ years of relevant professional experience
Experience in one of the following: Planning, Prediction, Reinforcement Learning, Imitation Learning, generative modeling (diffusion, autoregressive models)
Experience with training and deploying transformer-based model architectures
Experience with production Machine Learning pipelines: dataset creation, training frameworks, metrics pipelines
Fluency in Python ML frameworks and a basic understanding of C++
Bonus Qualifications
Top tier publications (NeurIPS, ICML, CVPR)
Experience with JAX
Similar jobs
Search more Machine Learning Engineer jobsSenior Machine Learning Engineer, Driving Product
🕒 8 days ago
Wayve👥 501 - 1000 employees🏢 Computer Software
🕒 8 days ago
Develop and ship autonomous-driving ML models for L2–L4 vehicles, owning data curation, training, evaluation, and on-road iteration to improve driving behaviours and real-world performance.
🕒 yesterday
Lead the development and deployment of production-grade machine learning systems, shaping architecture and engineering standards while advising clients and mentoring engineers.
Senior Machine Learning Engineer - Video Understanding
🕒 4 days ago
Scandit👥 201 - 500 employees🏢 Computer Software
🕒 4 days ago
Lead technical discovery and development of video-based computer vision systems for retail, building end-to-end ML pipelines and action-recognition models while validating performance and production constraints.
Senior Machine Learning Engineer - Video Understanding
🕒 4 days ago
Scandit👥 201 - 500 employees🏢 Computer Software
🕒 4 days ago
Lead the 0-to-1 development of video-based computer vision systems for retail, owning technical experiments, data workflows, model development and deployment to deliver robust action recognition in live environments.