About the Role
Our client is looking for a creative and driven Machine Learning Engineer to join our autonomous vehicle team, with a primary focus on advancing motion planning systems. In this role, you will help build the intelligence that enables vehicles to make safe, real-time decisions in complex, dynamic environments. You’ll work at the intersection of machine learning, robotics, and control systems - developing models that not only understand the world, but actively determine how the vehicle should move through it. If you’re excited about solving high-stakes planning challenges and shaping the decision-making layer of autonomous systems, we’d love to connect.
What You’ll Do
- Design, train, and deploy state-of-the-art machine learning models specifically for motion planning and decision-making in real-world driving scenarios
- Develop models that generate safe, efficient, and human-like trajectories by reasoning over dynamic agents, uncertainty, and long-term outcomes
- Build and maintain scalable data pipelines to process and learn from large-scale sensor and simulation datasets, with a focus on planning-relevant features
- Apply advanced deep learning architectures (e.g., transformers, sequence models) to capture temporal dependencies and predict multi-agent interactions that inform planning decisions
- Define and own planning-centric evaluation metrics, ensuring model performance aligns with safety, comfort, and real-world driving behavior
- Collaborate closely with software and systems engineers to integrate planning models into real-time, on-vehicle inference systems with strict latency constraints
- Explore and apply techniques from reinforcement learning, imitation learning, and optimization to improve planning robustness and adaptability
- Stay at the forefront of research in motion planning, decision-making under uncertainty, and autonomous systems
What You’ll Need:
- Strong proficiency in Python and hands-on experience with modern deep learning frameworks (e.g., PyTorch, TensorFlow, or JAX)
- Deep understanding of machine learning fundamentals, with particular interest in sequential modeling, decision-making, or control systems
- Experience across the full ML lifecycle, with exposure to deploying models in real-time or safety-critical environments
- Proficiency in C++ for implementing high-performance inference and planning systems
- Must be willing to work onsite 5 days a week in Austin, TX - local candidates preferred.
Nice to Have:
- Experience applying machine learning to motion planning, behavioral prediction, or robotics decision-making problems
- Familiarity with reinforcement learning, trajectory optimization, or control theory
- Experience with MLOps tools (e.g., MLflow, Kubeflow, Weights & Biases) and scalable training frameworks (e.g., Spark, Ray)
- Contributions to open-source ML projects or strong performance in competitions (e.g., Kaggle)
- Publications in top-tier ML or robotics conferences (e.g., NeurIPS, ICML, ICLR, CoRL, RSS, CVPR)
