Internship 2027 Behavior Machine Learning Engineer, World Models

Bedrock Robotics; San Francisco (CA); On-site; USD 45,000 - 60,000; Full time

Job summary

Bedrock Robotics is seeking an internship to advance autonomous construction AI. You will train world models on real fleet data, perform architecture explorations, and collaborate with Behavior and Controls teams to ship usable solutions for excavators.

The role emphasizes hands-on model development, evaluation, and rigorous experimentation on real-world projects, with opportunities to contribute to safety and efficiency on site.

Qualifications

• Pursuing BS, MS, or PhD in computer science, ML, robotics, or a related field—or equivalent research/industry experience.
• Strong Python and hands-on experience training models in PyTorch.
• Ability to read a paper and turn it into a working implementation.
• Comfort with messy, real-world data—not a curated benchmark.
• Sound experimental judgment distinguishing real results from noise.

Responsibilities

• Train world models on real fleet data from autonomous machines.
• Run architecture explorations and ablations and justify what works.
• Build evaluation harnesses and metrics to compare models.
• Collaborate with behavior and controls teams to integrate outputs into planning and control.
• Present findings and lay a solid foundation for the team's next steps.

Skills

Python

PyTorch

Reading papers

Experimental judgment

Education

BS/MS/PhD in CS/ML/Robotics or related field

Job description

Join the team bringing advanced autonomy to the built world

At Bedrock, we're moving AI out of the lab and into the real world. Our team includes veterans who helped launch Waymo, scaled Segment to a $3.2B acquisition, and grew Uber Freight to $5B in revenue. Today, we're deploying autonomous systems on heavy construction equipment across the country, improving safety on job sites and accelerating schedules on critical infrastructure projects.

We're not here debating the future of AI. We're deploying it in the real world. In just two years, we've raised $350M and achieved the first fully autonomous excavator deployments in construction.

This is where algorithms meet steel-toed boots. You'll work alongside construction veterans and world-class engineers to solve physical-world problems that simulations can't touch. If you're ready to do meaningful work on hard problems, we'd love to have you join us.

About the Role & Team

Predicting what happens next on a construction site is one of the hardest and most valuable problems at Bedrock, whether that's soil dynamics, machine interactions, or the people moving around the site. Every scoop of dirt we move rewrites the scene we have to plan against. The Behavior ML team is building world models that let our machines reason about the near future instead of just reacting to the present. As our intern, you'll train models on real fleet data, run architecture explorations, and work closely with the Behavior and Controls teams to ship what you build onto a real excavator.

What You'll Do

• Train and evaluate world models on real fleet data collected from autonomous machines in the field

• Run architecture explorations and ablations, and make a defensible case for what's actually working

• Build evaluation harnesses and metrics that tell us whether a model is genuinely better

• Partner with the behavior and controls teams to connect model outputs to planning and control

• Present your findings and provide a solid foundation for the team to build upon

What We're Looking For

Required

• Currently pursuing a BS, MS, or PhD in computer science, machine learning, robotics, or a related field - or bringing equivalent research or industry experience

• Strong Python and hands-on experience training models in PyTorch (or equivalent)

• Ability to read a paper and turn it into a working implementation

• Comfort with messy, real-world data - this is not a curated benchmark

• Sound experimental judgment understanding the difference between a real result and noise

Preferred

• Experience with world models, video prediction, or sequence modeling (e.g., DreamerV3, V-JEPA 2, Genie 3, GAIA-2, NVIDIA Cosmos, etc.)

• Background in reinforcement learning or imitation learning

• Prior work with robotics or autonomous vehicle data

• Familiarity with large-scale training infrastructure

Bedrock Robotics is an Equal Opportunity Employer

We're committed to building a diverse and inclusive workplace. We consider all qualified applicants for employment without regard to race, color, religion, sex, sexual orientation, gender identity, national origin, ancestry, age, disability, veteran status, genetic information, or any other protected characteristic.

Reasonable Accommodations

We want our hiring process to be accessible to everyone. If you need an accommodation to participate in the application or interview process, please let your recruiter know so we can support you.