Bot Auto
Algorithm Engineer, Autonomy Planning & Prediction
About this role
Design and deploy motion planning, prediction, and mapping algorithms for autonomous trucks operating on highways and surface streets. You'll own the full lifecycle from algorithm design through real-world validation, working with classical and learning-based approaches to enable safe autonomous freight transportation.
What you'll do
- Develop and optimize motion planning and behavior algorithms for autonomous highway and surface-street driving
- Build learning-based prediction and planning pipelines including data curation, model training, and vehicle integration
- Advance online mapping and map-building systems that feed real-time signals to the planning stack
- Analyze failures from road deployments and simulation to identify root causes and implement algorithmic fixes
- Collaborate across perception, control, simulation, and operations teams to validate and release features
What they're looking for
- C++ and Python programming
- Motion planning and optimization algorithms
- Prediction and behavior modeling
- Machine learning model design and training
- Online and HD mapping systems
- Real-time system optimization
- Autonomous vehicle software architecture
- Safety-critical systems validation
Opens the application — the Jobs AI extension fills it for you. Set up autofill
Opens the official application on the employer’s site. No login required.
Bot Auto
Bot Auto builds autonomous truck technology, developing the mechanical systems, deep learning models, and operational software that power self-driving commercial freight vehicles. The company is hiring mechanical engineers, machine learning engineers, software engineers, and interns to work across hardware design, perception and control systems, ML infrastructure, and fleet management platforms.
View all jobs at Bot AutoLikely interview questions
- Walk us through a motion planning or prediction algorithm you shipped in production—what were the key challenges and how did you validate it on real hardware?
- How have you approached integrating learning-based models into a real-time planning or prediction system, and what performance constraints did you have to work within?