ML Infra Engineer: GPU Fleet & Robotics

Generalist AI; Boston (MA); Vor Ort; USD 140.000 - 210.000; Vollzeit

Zusammenfassung

Generalist AI in Boston is building large robot foundation models, requiring massive GPU hardware and distributed training systems. You will own and optimize both training and inference infrastructure, including on‑prem GPUs connected to robots with strict real‑time latency budgets.

You will manage GPU compute fleets, ensure researchers can access GPUs easily, and optimize data loading and storage across distributed environments.

Qualifikationen

• Experience managing large GPU fleets for distributed training or inference.
• Deep experience with Slurm or Kubernetes for ML workloads.
• Experience building high-scale ML data loaders and pipelines.
• Strong understanding of ML hardware, storage and networking stacks.
• Experience working with NVIDIA GPU ecosystem and tooling.

Aufgaben

• Own our GPU compute fleets and ensure they are reliable and well‑documented.
• Make GPUs easy for researchers to use and maximize utilization.
• Optimize ML data loading, transport, and storage in highly distributed environments.
• Orchestrate robot inference fleets to meet real‑time latency requirements.

Kenntnisse

GPU fleet management

Slurm/Kubernetes

ML data loaders

Hardware/storage/networking know‑how

NVIDIA ecosystem

Tools

Slurm

Kubernetes

Jobbeschreibung

Generalist AI in Boston is building large robot foundation models, requiring massive GPU hardware and distributed training systems. You will own and optimize both training and inference infrastructure, including on‑prem GPUs connected to robots with strict real‑time latency budgets.

You will manage GPU compute fleets, ensure researchers can access GPUs easily, and optimize data loading and storage across distributed environments.