Research Engineer World Models
WaabiVisit Waabi website
Waabi is a Physical AI company developing autonomous-driving technology for commercial trucking and robotaxis.
Toronto, Canada
About Waabi
Founded in 2021 by Raquel Urtasun, Waabi builds a Physical AI platform combining a verifiable end-to-end AI model with neural simulation. Its Waabi Driver and Waabi World technologies support autonomous trucking and robotaxi development.
Skills
Candidate Availability
Required and preferred rules are kept separate and reflect the wording in the original posting.
About the Role
You will build generative and predictive world models for dynamic driving environments. You will develop distributed training and inference pipelines, optimize model deployment, build high-quality training data pipelines, maintain model infrastructure, and apply advances in generative AI and robotics.
Requirements
- Python
- PyTorch
- JAX
- Software engineering
- Distributed training
- Model deployment
- Generative model
- Predictive model
- World model
- Computer vision
- Machine learning
- Robotics
- Data pipeline
- Distributed system
Responsibilities
- Design, implement, and scale generative and predictive world-modeling systems
- Develop video, multimodal, and language-vision-action models for driving scenes
- Build generative scenario models and controllable simulation
- Translate model prototypes into distributed training and inference pipelines
- Optimize training and inference for large-scale datasets
- Build data pipelines for high-quality training datasets
- Maintain world-model codebase and infrastructure
- Apply advances in generative AI, distributed systems, and robotics deployment
Benefits
- Equity awards
- Annual performance bonus
- Medical coverage
- Dental coverage
- Vision coverage
- Unlimited vacation
- Flexible hours
- Work from home support
- Drinks, snacks, and catered meals when in office
- Team-building activities and social events