Forward Deployed Engineer - Physical AI

About Nebius: Nebius is leading a new era in cloud infrastructure for the global AI economy. We are building a full-stack AI cloud platform that supports developers and enterprises from data and model training through to production deployment, without the cost and complexity of building large in-house AI/ML infrastructure. Built by engineers, for engineers. From large-scale GPU orchestration to inference optimization, we own the hard problems across compute, storage, networking and applied AI. Listed on Nasdaq (NBIS) and headquartered in Amsterdam, we have a global footprint with R&D hubs across Europe, the UK, North America and Israel. Our team of 1,500+ includes hundreds of engineers with deep expertise across hardware, software and AI R&D. The role Nebius is building the cloud infrastructure that will power the next generation of Physical AI: robotics, autonomous systems, simulation, world models, and embodied intelligence operating in the real world.   The Forward Deployed Engineer, Physical AI Systems is a senior, high-autonomy individual contributor role that owns the technical bridge between customer data, AI models, simulation workflows, evaluation systems, and real-world deployment feedback. This role sits with strategic customers and ISV partners, embedded directly inside their engineering teams, and ships production software that turns messy real-world physical AI problems into reliable, measurable platform workflows.   You will work alongside the Field CTO and the Head of Physical AI. Inside each account, you own end-to-end technical execution: discovery, scoping, model and evaluation pipeline design, build, and production rollout. Across accounts, you identify the patterns worth productizing and then partner with Product and Engineering to fold them into the core platform. Your field work is the primary input to the Nebius Physical AI roadmap, and your job is to prove the platform's core claim: that customers can move from real-world failures to measurable model improvement faster than their internal tools allow, repeatably and at scale.   We are looking for engineers with the seniority and judgment of a founding engineer or staff individual contributor, people who can show up at a robotics company or a world model lab on a Monday and have credibility with the CTO by Friday. You will be trusted to make consequential technical decisions in ambiguous environments without waiting for permission.   In return, you get founder-level autonomy in an IC seat, direct exposure to the most important companies defining Physical AI, and the resources of a public cloud platform behind every line of code you ship. This is a definitive zero-to-one opportunity to write the code that defines the highest-growth segment of AI. You are welcome to work remotely from the United States (SF Bay Area, CA or Austin, TX preferred).   Your responsibilities will include:   End-to-End Ownership Inside Strategic Accounts: Own discovery, technical scoping, system design, build, and production rollout for each design partner and ISV engagement. Partner directly with customer engineering and domain teams to translate ambiguous problems into deployable production systems.  Physical AI Workflows & Pipelines: Build and own physical AI workflows across real-world data, synthetic data, model training, evaluation, deployment, and failure capture. Develop practical ML and evaluation pipelines for perception, autonomy, world models, and policy-learning use cases, operating inside the customer's codebase, on their infrastructure, against their data.  Evaluation & Failure Loops: Design scenario-based evaluation workflows, regression testing, failure analysis, and before-versus-after model comparisons. Help define the real-to-sim-to-real and failure-to-retrain loops, and convert customer failures into product insights, datasets, scenarios, tests, and retraining loops.  Customer Data & Integration: Work with customer datasets including video, images, telemetry, annotations, simulation outputs, and deployment logs. Integrate NVIDIA ecosystem tools where useful, including Isaac Sim, Isaac Lab, Cosmos, NeMo, GR00T, and Jetson, and decide where the platform should build, buy, or integrate across labeling, synthetic data, simulation, training, evaluation, and monitoring.  ISV Integration Development: Stand up custom technical integrations with key Physical AI ecosystem partners (simulation frameworks, robotics toolchains, data management vendors). Build the reference architectures and joint solutions that turn ISV partnerships into deployable, repeatable assets.  Pattern Codification & Productization: Identify which prototypes contain generalizable abstractions worth hardening into modular product components. Partner with the Field CTO, Product, and Engineering teams to fold these into the core Physical AI platform. Treat every engagement as a forcing function for the next ten.  Rapid Engineering V