OpenAI
Software Engineer, Compute Foundations
$255K–$490K
San Francisco · Full-time · Posted Sep 16, 2026
About the job
Build and operate distributed software that provisions, manages, and recovers GPU compute infrastructure across sites, data centers, and providers.
What you'll do
- Develop Kubernetes-based controllers, distributed services, APIs, and resource models for infrastructure lifecycle management.
- Build provisioning, configuration, upgrade, maintenance, recovery, and decommissioning workflows for machines and clusters.
- Improve reliability, throughput, latency, and recovery across hardware, operating-system, service, and provider boundaries.
What they ask for
- Strong software engineering fundamentals and experience designing, implementing, and owning production distributed systems or infrastructure services.
- Experience developing infrastructure systems using Kubernetes APIs and reconciliation to manage resources.
- Understanding of bare-metal provisioning, with depth in at least one area such as PXE, DHCP/DNS, BMCs, firmware, Linux, drivers, images, or configuration management.
- Ability to design reliable APIs and asynchronous workflows and reason about concurrency, consistency, idempotency, and failures.
Nice to have
- Built infrastructure control planes operating across multiple sites or regions.
- Worked with GPU or HPC infrastructure, or integrated multiple hardware platforms or infrastructure providers.
Pay
$255K – $490K USD base salary
Tools
Kubernetes, PXE, DHCP/DNS, BMCs, firmware, Linux
Model-written summary · the original posting is authoritative