Forward Deployed Engineer, Physical AI Infrastructure

Nebius is hiring a Forward Deployed Engineer to work with Physical AI customers on training, simulation, synthetic data, evaluation, and model serving workloads. The role combines customer engineering with product discovery: you will turn production requirements and measured customer needs into input for Nebius Cloud, Serverless, and Token Factory teams, while assessing new product opportunities and testing prototypes with customers.

The work is hands-on across Python, Linux, Kubernetes, GPUs, networking, storage, and model servers. You will build agents to set up and debug workloads, deploy and benchmark open-weight models using tools such as vLLM or SGLang, and tune systems against customer requirements. Nebius seeks at least three years in software engineering or infrastructure work, including ownership of a production service, cluster, or platform, plus practical experience with LLMs or agents and direct customer.

Highlights

• Translate Physical AI workloads into measurable infrastructure and product requirements.
• Build agents and prototypes, and evaluate performance, reliability, and product fit.
• Deploy and tune models with serving stacks such as vLLM or SGLang.
• Work across Kubernetes, GPUs, networking, storage, and customer engineering.