Staff Engineer, GPU Inference & Training Platform

United States Digital Space LLC; San Francisco (CA); Hybrid; USD 180,000 - 240,000; Full time

Job summary

United States Digital Space LLC is seeking an infrastructure leader to own a self-serve GPU compute platform for training and inference workloads. You will design and operate the system that lets researchers launch jobs across multi-cloud GPU fleets without manual provisioning.

You will own provisioning, scheduling, and reliability across providers, building fault-tolerance, observability, and a coherent platform roadmap for scalable AI workloads.

Qualifications

• Deep Kubernetes experience with custom operators and CRDs.
• Experience managing GPU clusters at scale with NVIDIA GPUs and high-speed networks.
• Experience orchestrating compute across multiple clouds.
• Strong distributed systems fundamentals.
• Proficiency in Go, Rust or C++ for infrastructure.
• Experience with long-running training and high-availability inference.
• Ability to own problems end-to-end.

Responsibilities

• Build a self-serve compute platform for training and inference workloads.
• Operate the GPU fleet across providers with provisioning and lifecycle management.
• Develop scheduling and placement to balance capacity across clouds.
• Support both long-running training jobs and production inference workloads.
• Own Kubernetes orchestration across multiple clusters and providers.
• Implement fault tolerance, autoscaling, and observability.
• Set technical direction across teams to align platform roadmap.

Skills

Kubernetes

GPU clusters

Multi-cloud orchestration

Distributed systems

Go/Rust/C++

Tools

Custom Kubernetes Operators

Job description

United States Digital Space LLC is seeking an infrastructure leader to own a self-serve GPU compute platform for training and inference workloads. You will design and operate the system that lets researchers launch jobs across multi-cloud GPU fleets without manual provisioning.

You will own provisioning, scheduling, and reliability across providers, building fault-tolerance, observability, and a coherent platform roadmap for scalable AI workloads.