Software Engineer, Compute Foundations Systems

About this role

ABOUT THE TEAM

Frontier Systems Foundations, part of Compute Foundations at OpenAI, builds the systems software foundation that turns new compute infrastructure into reliable, usable capacity for frontier model training.

Our mission is to make some of the world's largest GPU clusters work reliably for frontier training. We bring new platforms and clusters online, safely maintain installed fleets, and partner with hardware, infrastructure, and research teams to resolve the system-level issues that keep jobs from running.

That means building and maintaining the software closest to the machine: Linux and Ubuntu operating-system images, kernels and modules, drivers, packages and repositories, disks and boot configuration, firmware integration, provisioning, and system-level validation. We make these components reproducible, compatible, and safe to operate across heterogeneous fleets.

ABOUT THE ROLE

We are looking for systems software engineers with deep Linux and host-systems experience to build, qualify, and maintain the operating-system foundation for OpenAI's frontier compute fleet. Relevant backgrounds include kernel and module development, Linux distribution or image engineering, package management, firmware and driver integration, disks and boot, and bare-metal provisioning.

You'll work closely with hardware engineers, vendors, and infrastructure teams to bring up new platforms, integrate system components, and debug failures across firmware, disks, boot, operating systems, kernels, drivers, and workload interactions. Your work will directly influence how quickly new capacity becomes usable and how reliably large GPU fleets operate.

You should be comfortable writing and maintaining production-quality systems software and automation, but we do not expect expertise across every layer. This is an opportunity to go deep on challenging systems problems while building the image, package, qualification, and recovery paths that power the next generation of frontier models.

IN THIS ROLE, YOU WILL:

- Build and maintain the Linux host software stack for large GPU clusters, including Ubuntu and OS images, kernel configuration and modules, drivers, packages, disks and storage configuration, and machine configuration.

- Design reproducible OS-image builds, package and repository workflows, and system configuration for heterogeneous bare-metal and cloud compute fleets.

- Integrate, test, and qualify kernels, modules, drivers, packages, and firmware across new and existing hardware platforms; build safe canary, rollback, and recovery paths for system changes.

- Bring up new hardware platforms and compute SKUs, working with hardware engineers and vendors to resolve firmware, driver, operating-system, and compatibility issues.

- Debug complex system failures across boot and provisioning, firmware, disks, kernels and drivers, and workload interactions; turn recurring failure modes into durable fixes, tests, and automation.

- Improve provisioning, repave, maintenance, and recovery correctness by eliminating manual host-by-host intervention and making system behavior predictable at fleet scale.

- Build focused systems tooling and diagnostics that make the host software stack easier to validate, troubleshoot, and operate across the installed fleet.

YOU MIGHT THRIVE IN THIS ROLE IF YOU:

- Have significant experience building, integrating, or operating production Linux systems, particularly Ubuntu, Debian, or another major Linux distribution.

- Have deep experience in one or more of: Linux kernels, modules, or device drivers; Linux distribution, package, repository, or OS-image engineering; boot, provisioning, disks, or machine configuration; firmware and driver integration, qualification, or rollout.

- Can write, debug, and maintain production-quality systems software and automation using appropriate systems languages, scripting, and Linux tooling.

- Understand how to build, test, package, qualify, and safely deliver system-level changes, including compatibility testing, staged rollout, rollback, and recovery.

- Can systematically debug failures that cross hardware, firmware, boot, operating-system, kernel, and driver boundaries, and collaborate effectively with hardware and software partners.

- Enjoy investigating difficult system behavior, identifying underlying failure modes, and building pragmatic fixes that make production systems more reliable.

BONUS POINTS IF YOU:

- Have built bare-metal provisioning, machine-bootstrap, OS-reprovisioning, or system-lifecycle tooling.

- Have worked with GPU systems, AI infrastructure, HPC clusters, accelerators, or other heterogeneous and performance-sensitive compute environments.

- Have experience bringing up new server platforms or working directly with hardware and software vendors on firmware, driver, or operating-system compatibility issues.

- Have contributed to Linux, kernel subsystems or modules, distributions, package ecosystems, drivers, firmware tooling, or other open-source systems software.

- Have delivered or supported system-software changes across large production compute fleets.

About OpenAI

OpenAI is an AI research and deployment company dedicated to ensuring that general-purpose artificial intelligence benefits all of humanity. We push the boundaries of the capabilities of AI systems and seek to safely deploy them to the world through our products. AI is an extremely powerful tool that must be created with safety and human needs at its core, and to achieve our mission, we must encompass and value the many different perspectives, voices, and experiences that form the full spectrum of humanity. 

We are an equal opportunity employer, and we do not discriminate on the basis of race, religion, color, national origin, sex, sexual orientation, age, veteran status, disability, genetic information, or other applicable legally protected characteristic.

For additional information, plea

Similar jobs

• Software Engineer, Compute Platform — Replit · Foster City, CA
• Software Engineer, Compute Infrastructure — Openai · San Francisco
• Software Engineer, Compute Infrastructure — Render · Remote
• Software Engineer, Compute — Persona · San Francisco
• Software Engineer, Compute — Vercel · Remote
• Software Engineer, Compute Team — Planetlabs · Remote
• Senior Software Engineer, Compute Platform — Reddit · Remote
• Software Engineer, Foundations — Hextechnologies · Remote