OpenAI
Software Engineer, Compute Foundations Systems
$230K–$490K
San Francisco · Full-time · Posted Jul 21, 2026
About the job
Builds and maintains the Linux host software foundation that brings GPU compute platforms online and keeps large fleets reliable for frontier model training.
What you'll do
- Build and maintain Linux host software, including OS images, kernels, drivers, packages, storage, and machine configuration.
- Design reproducible image-build and package workflows, and qualify system components with safe rollout, rollback, and recovery paths.
- Bring up hardware platforms and debug cross-layer system failures, turning recurring issues into durable fixes, tests, and automation.
What they ask for
- Significant experience building, integrating, or operating production Linux systems.
- Deep experience in at least one area such as kernels and drivers, OS-image or package engineering, provisioning and disks, or firmware integration.
- Ability to write, debug, and maintain production-quality systems software and automation.
- Ability to qualify and safely deliver system changes and debug failures across hardware and software boundaries.
Nice to have
- Experience with bare-metal provisioning or system-lifecycle tooling.
- Experience with GPU, AI infrastructure, HPC, accelerators, heterogeneous compute, or large production fleets.
Pay
$230K – $490K USD base salary
Tools
Linux, Ubuntu, Debian
AI focus
Builds and maintains compute infrastructure used for frontier model training; no direct AI model development is described.
Model-written summary · the original posting is authoritative