2027 Internship Behavior ML Engineer, Learned Manipulation Policies

We Love AI Jobs; San Francisco (CA); On-site; USD 34,000 - 55,000; Full time

Job summary

Bedrock Robotics in San Francisco offers an on-site internship in Behavior ML focused on training learned policies for autonomous excavators using real demonstrations and simulations. You’ll contribute to the team’s training codebase and help evaluate policy performance across tasks like juggling safety and efficiency.

Ideal candidates are pursuing BS/MS/PhD in CS/robotics/ML, with strong Python and PyTorch experience, and familiarity with diffusion models, flow matching, VAEs, or imitation

Qualifications

• Currently pursuing BS, MS, or PhD in CS/robotics/ML or equivalent
• Strong Python and PyTorch experience training models
• Experience with diffusion models, flow matching, VAEs, or imitation learning
• Experience running training experiments and distinguishing real improvements from noise
• Good communication of results and next steps

Responsibilities

• Train and iterate on diffusion flow matching, and related behavior policies using real fleet demonstration data and simulated rollouts
• Run architecture, conditioning, and hyperparameter explorations, and turn results into clear findings
• Build and improve evaluation pipelines that score behavior models on task success, smoothness, safety margins, and operator-likeness
• Investigate failure modes - distribution shift, mode collapse, out-of-distribution scenes and propose fixes
• Work with the simulation and eval teams to ensure offline metrics predict on-machine performance
• Contribute to the shared training codebase with clean, reproducible work
• Share results regularly with the behavior, controls, and autonomy teams

Skills

Python

PyTorch

diffusion models

flow matching

imitation learning

behavior cloning

model evaluation

robot manipulation

Education

BS/MS/PhD in CS/Robotics/ML (in progress)

Job description

# 2027 Internship Behavior ML Engineer, Learned Manipulation PoliciesBedrock RoboticsSan Francisco, CAInternshipAI tools:PyTorchdiffusion modelsflow matchingVAEsVision Language Action (VLA) modelsUpload your resume once and Scout applies for youApply with ScoutShare this job:LinkedInXSave imageYou’ll train learned behavior policies that teach autonomous excavators to perform tasks such as swinging, digging, and dumping. The internship combines experiments with real demonstrations and simulated rollouts, evaluation of policy performance, and contributions to the team’s training codebase.* Internship* On-site* Entry Level* BS, MS, or PhD (in progress)### Skills and expertisePythonPyTorchdiffusion modelsflow matchingimitation learningbehavior cloningmodel evaluationrobot manipulation### Key responsibilities* Train and iterate on behavior policies using fleet demonstrations and simulated rollouts.* Evaluate models for task success, smoothness, safety margins, and operator-likeness.* Investigate policy failure modes and share reproducible findings with autonomy teams.## How this job comparesCompared with 45 open intern machine learning jobs on October 10, 2026.Similar open jobs: We Love AI Jobs lists 45 open intern machine learning jobs at 26 companies.Remote work: This job is on site. 2 of the 45 intern machine learning jobs (4%) are remote.AI in this role: High. AI is the core of the work: building, training or deploying AI systems. Training diffusion and related behavior models is the core of the role. The intern will also build evaluation pipelines and investigate model failure modes.## Full descriptionJoin the team bringing advanced autonomy to the built worldAt 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 & TeamTeaching a 40-ton excavator to move like an expert operator is a very different problem from teaching a car to stay in its lane. The motions are multimodal, there are many good ways to swing, dig, and dump and the consequences of getting them wrong are measured in cubic yards and bent steel. Bedrock's Behavior ML team builds the learned policies that decide what our machines actually do, and diffusion policies are a central bet: models that can represent the full distribution of expert behavior instead of averaging it into mush. As our Behavior ML intern, you'll help train those policies and build the evaluation that tells us whether they're genuinely better working alongside two hosts who sit on both the training and evaluation sides of the problem.What You'll Do• Train and iterate on diffusion flow matching, and related behavior policies using real fleet demonstration data and simulated rollouts• Run architecture, conditioning, and hyperparameter explorations, and turn the results into clear findings the team can build on• Build and improve evaluation pipelines that score behavior models on task success, smoothness, safety margins, and operator-likeness• Investigate failure modes - distribution shift, mode collapse, out-of-distribution scenes and propose fixes• Work with the simulation and eval teams to make sure offline metrics actually predict on-machine performance• Contribute to the shared training codebase with clean, reviewed, reproducible work• Share results regularly with the behavior, controls, and autonomy teamsWhat We're Looking ForRequired• Currently pursuing a BS, MS, or PhD in computer science, robotics, machine learning, or a related field or bringing equivalent research or industry experience• Strong Python and hands-on experience training models in PyTorch (or equivalent)• Working understanding of generative modeling diffusion models, flow matching, VAEs, or similar and of imitation learning or behavior cloning• Experience running and interpreting real training experiments: you know how to tell a real improvement from noise• Clear communication, you can explain what you tried, what happened, and what you'd do nextPreferred• Published work or substantial project experience in diffusion, flow matching, robot learning, or imitation learning• Experience training and building Vision Language Action (VLA) models• Experience training Reinforcement Learning (RL) policies for robot manipulation• Experience evaluating policies in simulation and reasoning about the sim-to-real gap• Familiarity with large-scale training infrastructure and experiment tracking tooling• Exposure to robotics, autonomous vehicles, or other physical-world control problems• Interest in construction, earthwork, or heavy equipment - no prior experience requiredBedrock Robotics is an Equal Opportunity EmployerWe’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.