Torc Robotics is a leader in autonomous driving technology, focused on developing software for automated trucks. The Senior Machine Learning Engineer will develop and scale learning-based systems to enhance driving behavior and ensure safe, efficient autonomy for freight operations.
Responsibilities:
- Own development and delivery of End-to-End ML models that map multi-modal sensor inputs (camera, LiDAR, radar, maps) to driving-relevant outputs (trajectories, cost functions, or intermediate representations)
- Train and evaluate models using large-scale datasets from fleet logs, simulation, and synthetic data
- Analyze model performance, identify failure modes, and drive data-driven improvements in robustness and generalization
- Design and refine training pipelines, data workflows, and evaluation strategies to improve iteration speed and model quality
- Contribute to model architecture decisions, including approaches such as imitation learning, reinforcement learning, transformers, and vision-language-action (VLA) models
- Collaborate closely with Perception, Prediction, Planning, and Simulation teams to ensure alignment across the autonomy stack
- Support integration of E2E models into simulation and on-vehicle systems for closed-loop validation
- Improve tooling, experimentation workflows, and reproducibility across the team
- Mentor junior engineers and contribute to team-level best practices and technical discussions
Requirements:
- Bachelor's degree with 6+ years, Master's with 4+ years, or PhD with 0–2 years of experience in Machine Learning, Robotics, Computer Science, or a related field with a track record of publications in top-tier conferences (e.g., NeurIPS, ICML, ICLR, CVPR, ICCV, CoRL)
- Experience developing and deploying ML models for autonomous systems, robotics, or complex decision-making environments
- Strong programming skills in Python and PyTorch, with ability to write production-quality ML code
- Experience training and evaluating models using large-scale datasets and distributed compute environments
- Solid understanding of ML architectures used in E2E systems, such as Transformers, BEV models, VLA/VLM approaches, or diffusion models
- Proven ability to debug model behavior, analyze performance metrics, and drive iterative improvements
- Experience contributing to or influencing model architecture and training strategies
- Ability to work cross-functionally and integrate ML systems into larger autonomy pipelines
- Experience developing End-to-End or mid-to-end models for autonomous driving or robotics
- Experience with vision-language models (VLMs) or vision-language-action (VLA) systems
- Familiarity with closed-loop simulation and evaluation frameworks
- Experience with reinforcement learning or imitation learning in real-world systems
- Experience with distributed training frameworks (e.g., Ray)
- Understanding of vehicle dynamics, motion planning, or multi-agent systems