3 months ago
Responsibilities
- Develop and train diverse conditioned policies that simulate realistic driving behaviors for stress-testing and validating autonomous-driving systems.
- Lead research and implementation of reinforcement-learning algorithms with safety metrics as primary learning constraints.
- Design reward functions and evaluation metrics balancing safety, progress, and comfort.
- Optimize large-scale, high-throughput training environments for complex multi-agent scenarios.
- Advance neural architectures for spatial reasoning, long-horizon planning, and interaction modeling.
- Collaborate with Simulation and Planning teams to integrate research-grade models into production-quality safety-critical software.
Requirements
- Proven experience training and deploying deep reinforcement-learning algorithms such as PPO and SAC for complex real-world robotic or autonomous systems.
- Expertise in Python and PyTorch, with a strong understanding of modern deep-learning architectures and optimization techniques.
- MS or PhD in Computer Science, Robotics, or a related quantitative field.
- Ability to diagnose and solve reinforcement-learning challenges including variance management and distribution shift.
- Preferred experience with constrained optimization or safety-critical learning frameworks.
- Preferred background in multi-agent reinforcement-learning stability, self-play, and decentralized execution.
- Familiarity with vehicle dynamics and behavior planning for long-haul highway environments is preferred.
Benefits
- Comprehensive health insurance and paid time off.
- Performance bonuses and equity opportunities are mentioned.
- Opportunity to work on autonomous trucking and safety-critical AI systems.
Categories
ML EngineeringRobotics
About Bot Auto
Bot Auto runs autonomous trucks out of Houston, Texas. We sell the capacity, not the technology. We own the trucks, the safety case, and the operation end to end. Our team pairs trucking operators who've run real fleets with the engineers building the technology because commercializing this industry takes both. We're not chasing a demo. We're building the freight network the industry actually needs.