robot learning research engineer (manipulation policies)

Start Date: ASAPAbout UsMundane is a venture-backed seed-stage robot learning startup founded by a team of Stanford researchers and builders. We’re deploying a massive fleet of humanoid robots to perform mundane tasks in commercial environments, collecting data to build the next generation of embodied intelligence.We are a fast-paced, execution-driven team of engineers, roboticists, and builders. Our robots operate in real customer environments — and improve through real-world experience.About the RoleYou will develop and ship learning-based manipulation policies that run on real robots.Our robots already collect real-world data and execute manipulation tasks. Your role is to turn that data into policies that improve reliability, generalize across tasks, and hold up under real-world distribution shift.This role is deeply hands-on and execution-focused. You will implement models, run controlled experiments, and validate improvements directly on physical robots. Success is measured by real-world performance, not benchmark metrics.At Mundane, your models will not live in simulation or papers — they will deploy to humanoid robots operating in customer environments.What You’ll OwnDevelopment and improvement of real-world manipulation policiesPolicy architecture and training recipes for real-world manipulationRobustness improvements (recovery behaviors, partial observability, drift, edge cases)Experiment discipline and clear ablation methodologyScaling from single-task policies to multitask robot capabilitiesPackaging models for deployment on real robotsResponsibilitiesExtend and improve our policy learning stack (imitation learning / sequence-based policies) for real-world manipulation tasksDesign and run disciplined experiments to improve policy performance, including clear ablations and controlled comparisonsDevelop multitask policies with effective task conditioning and thoughtful data mixture strategiesImprove robustness through techniques such as data augmentation, recovery behaviors, and training under partial observabilityDesign and run systematic stress tests to evaluate distribution shift, drift, and edge-case failuresWork closely with infrastructure engineers to scale training pipelines and experiment workflowsCollaborate with reliability engineers to define evaluation gates and deployment criteriaPackage trained models for deployment, addressing latency, stability, and safety constraintsInvestigate real-world failures and iterate rapidly to improve policy robustnessQualificationsStrong PyTorch and ML engineering skills with the ability to implement and ship reliable training pipelinesPractical experience with imitation learning or behavior cloningExperience training sequence-based models such as transformers, diffusion policies, or related architecturesComfort running real-world experiments and debugging issues across data, training, and deploymentStrong experimental rigor, including designing ablations, maintaining reproducibility, and avoiding “demo-only” improvementsNice to HaveExperience with robotic manipulation systems and real-world robot experimentationFamiliarity with common failure modes in manipulation tasksExperience scaling training across large datasets or multi-GPU environmentsBackground in embodied AI or robot learning systemsWhat You’ll GetDirect ownership over the policies that control robots operating in real environmentsEarly equity with meaningful upside in a venture-backed robotics companyThe opportunity to see your research deployed quickly on real humanoid robotsClose collaboration with hardware, infrastructure, and deployment teamsA front-row seat in scaling a technically ambitious robotics company from seed stagePerks: Competitive salary + equity, flexible PTO, legendary merch, coffee, robots, sauna & cold plunge (pending)