Machine Learning Engineer, Motion Planning, Self-Driving

About the role

You will own the design and implementation of motion planning algorithms that power the decision-making stack for autonomous vehicles. Your work will focus on building and optimizing planning models that operate in complex and dynamic real-world driving scenarios. You will collaborate closely with perception and control teams to ensure end-to-end system robustness and safety. You will translate high-level autonomy requirements into concrete model architectures and training objectives. You will analyze fleet data to identify edge cases and improve the generalization of planning behaviors. You will iterate on simulation and on-road evaluation metrics to validate planning performance. Most importantly, you will see your algorithms deployed at scale and directly influence the safety and reliability of millions of customers.

Key facts

What you'll do

Design and implement motion planning algorithms that operate on high-dimensional state spaces and support real-time decision-making in autonomous driving systems.

Leverage large-scale driving data, including sensor streams and human interventions, to train models that robustly handle diverse traffic scenarios and edge cases.

Apply generative modeling, imitation learning, and reinforcement learning techniques to improve trajectory prediction, decision-making, and behavior planning.

Develop data generation pipelines and fleet learning strategies that continuously expand the diversity and quality of training data for planning models.

Build simulation environments and evaluation frameworks to quantify planning performance across safety metrics, efficiency, and passenger comfort.

Integrate planning components with vehicle firmware and control stacks to enable safe and reliable closed-loop operation in production vehicles.

Collaborate with perception and control engineers to align model interfaces, timing constraints, and safety requirements across the autonomy stack.

Translate high-level autonomy objectives into concrete model architectures, loss functions, and training procedures that meet strict reliability standards.

Monitor in-vehicle performance using telemetry and feedback loops to drive iterative improvements and rapid bug resolution.

Ship production-quality software that meets automotive safety standards and can scale across the entire Tesla vehicle fleet.

Contribute to technical documentation, code reviews, and cross-team design discussions to maintain high engineering standards.

Experiment with new planning paradigms and model representations to push the boundaries of what the driving models can achieve.

Support debugging and analysis of on-road incidents using data-driven tools and simulation-based reconstructions.

Mentor and guide junior engineers by sharing best practices in machine learning, software engineering, and autonomous systems.

Champion best practices for reproducibility, testing, and monitoring of machine learning models in safety-critical environments.

Requirements

Must have a Bachelor's degree in Computer Science, Electrical Engineering, or a related technical field, or equivalent practical experience.

Must have proven experience in machine learning or robotics, with a strong track record of deploying models in production environments.

Must have deep expertise in Python and experience with major deep learning frameworks such as TensorFlow or PyTorch.

Must have solid software engineering skills, including experience with version control, testing, and scalable code design.

Must be comfortable working with C++ to integrate algorithms with low-level vehicle firmware and real-time systems.

Must understand modern deep learning architectures, optimization techniques, and model alignment methods relevant to perception and planning.

Must have experience handling large-scale datasets and building data pipelines that support high-quality model training.

Must be able to work effectively in a fast-paced environment with ambiguous problems and evolving requirements.

Must have strong analytical skills to diagnose model behavior and performance issues using simulation and real-world data.

Must be willing to travel as required for team collaboration, project reviews, or on-road testing activities.

Must be eligible to work in the United States without sponsorship for this position.

Must be available to start within the required timeframe outlined during the hiring process.

Nice to have

Preferred experience in robotics, autonomous vehicles, or large-scale deployment of ML models in safety-critical systems.

Familiarity with motion planning libraries, control theory, and simulation tools used in autonomous driving.

Experience with optimization-based planning methods or classical control techniques alongside machine learning approaches.

Background in handling edge cases, rare events, and failure mode analysis in deployed autonomous systems.

Practical notes

This is a full-time position based in Austin.

Travel may be required as needed for team collaboration and on-road testing.

Candidates must be eligible to work in the United States without sponsorship.

Please adhere to the start date outlined during the hiring process if selected.