Job description

Who we are Aurora’s mission is to deliver the benefits of self-driving technology safely, quickly, and broadly. The Aurora Driver aurora.tech will create a new era in mobility and logistics, one that will bring a safer, more efficient, and more accessible future to everyone. At Aurora, you will tackle massively complex problems alongside other passionate, intelligent individuals, growing as an expert while expanding your knowledge. For the latest news from Aurora, visit aurora.tech aurora.tech or follow us on LinkedIn linkedin.com. What we are looking for We're searching for Summer 2027 Behavior & Planning Interns. You'll work on how the Aurora Driver decides what to do — forecasting how other agents move, planning and selecting trajectories, and the imitation-learning/reinforcement-learning models behind motion planning. In this role you will - Develop end-to-end planning or prediction models for the Aurora Driver using imitation learning, reinforcement learning, and/or classical motion planning - Formulate ambiguous driving-behavior problems as concrete ML or optimization problems - Reason about competing objectives (e.g. progress vs. safety) and uncertainty in your model design and evaluation - Work with autonomy/robotics-scale datasets and benchmarks, not just toy datasets - Present a project plan and regular progress updates to the team, working closely with a mentor Required qualifications - Currently pursuing a Bachelors, Master's or PhD in Computer Science, Robotics, or a related field, with a research or ML focus - Strong ML foundations with hands-on experience in PyTorch, including custom dataloaders, architectures, and loss functions - Some exposure to imitation learning, reinforcement learning, or motion planning - Experience training models on large-scale, non-toy datasets or benchmarks - Ability to reason mathematically about optimization objectives - Awareness of the compounding-error / distribution-shift problem in imitation learning Desirable qualifications - Publication record at a peer-reviewed venue - Experience with constrained optimization or preference tuning (e.g. RLHF, DPO) - Experience with online reinforcement learning or interactive learning (e.g. DAgger) Compensation The hourly pay rate for this internship in Mountain View or San Francisco is: - $52.00 per hour for interns pursuing a bachelor’s degree; - $58.00 per hour for interns pursuing a master’s degree; and - $65.00 per hour for interns pursuing a doctoral degree. The hourly pay rate for this internship in Pittsburgh is: - $47.00 per hour for interns pursuing a bachelor’s degree; - $52.00 per hour for interns pursuing a master’s degree; and - $58.00 per hour for interns pursuing a doctoral degree. Working at Aurora At Aurora, we bring together extraordinarily talented and experienced people united by the strength of our values. We operate with integrity, set outrageous goals, and build a culture where we win together — all without any jerks. We believe in-person work increases collaboration, empathy, and our ability to lead effectively. As a result, Aurorans are expected to work from the office at least 3 days per week, with some roles requiring more frequent in-office presence, up to 5 days per week, based on operational needs. Our Careers page aurora.tech provides insight into what it is like to work at Aurora, and you can find all the latest updates in our Newsroom aurora.tech. Our commitment to safety At the core of everything we do is our commitment to safety. Building best-in-class self-driving technology will take time, and we believe that each employee at Aurora has a role in contributing to safety, every step of the way. Aurora expects commitment to our safety policies from every employee, and seeks candidates who take an active responsibility, can contribute to building an atmosphere of trust, and invest in the organization’s long-term success by prioritizing working safely, no matter what. Our commitment to inclusion Aurora considers candidates without regard to their race, color, religion, national origin, age, sex, gender, gender identity, gender expression, sexual orientation, marital status, pregnancy status, parent or caregiver status, ancestry, political affiliation, veteran and/or military status, physical or mental disability, or any other status protected by federal or state law. Aurora considers qualified applicants with criminal histories, consistent with applicable federal, state, and local law. We are also committed to providing reasonable accommodations for qualified individuals with disabilities and disabled veterans in our job application procedures. If you need assistance or an accommodation due to a disability, you may contact us at careersiteaccommodations@aurora.tech. For California applicants, information collected and processed as part of your application and any job applications you choose to submit is subject to Aurora’s California Employment Privacy Policy.

Requirements

Must-haves

Currently pursuing a Bachelors, Master's or PhD in Computer Science, Robotics, or a related field, with a research or ML focus
Strong ML foundations with hands-on experience in PyTorch, including custom dataloaders, architectures, and loss functions
Some exposure to imitation learning, reinforcement learning, or motion planning
Experience training models on large-scale, non-toy datasets or benchmarks
Ability to reason mathematically about optimization objectives
Awareness of the compounding-error / distribution-shift problem in imitation learning
Desirable qualifications
Publication record at a peer-reviewed venue
Experience with constrained optimization or preference tuning (e.g. RLHF, DPO)
Experience with online reinforcement learning or interactive learning (e.g. DAgger)

Nice-to-haves

None flagged.

Application prep

Use the role’s language and application details while tailoring your materials.

Keywords to mirror

Platform: ashby
Effort: quick