About the Team

Cloud Infrastructure builds the shared platforms behind ChatGPT and the OpenAI API. We work across engineering and product teams to make infrastructure reliable, secure and easier to use as OpenAI grows.

Within Cloud Infrastructure, Kubernetes Foundations builds and operates the multi-cloud compute platform that other teams build on. Our work spans cluster creation and lifecycle management, upgrades, patching, compliance and the tools engineers use to provision and run infrastructure. The platform supports a broad range of customers, including API and chat products, data systems, observability and online serving.

Our focus is the shared CPU compute foundation and its platform capabilities, rather than specialized GPU optimization. We build across OCI, AWS and Azure, solving the practical problems of making a common platform useful across different cloud environments.

OpenAI develops and deploys AI with the goal of benefiting humanity. Responsible deployment and safety are central to that work.

About the Role

You will build infrastructure abstractions that help teams ship and scale products. This is a software engineering role with end-to-end production ownership: designing capabilities, writing the implementation, rolling it out and improving how it operates.

We are looking for engineers who have built shared infrastructure, not only deployed applications on an existing platform. You should be comfortable understanding users’ needs, identifying missing capabilities and turning an ambiguous problem into a working system. Experience at OpenAI’s scale is not a prerequisite; sound engineering judgment and the ability to adapt familiar solutions to a much larger environment matter.

This role is based in San Francisco, CA.

This role is exclusively based in our San Francisco HQ. We offer relocation assistance to new employees.

In this role, you will

Build and improve the systems that create, configure and manage Kubernetes clusters throughout their lifecycle, including safe upgrades and ongoing maintenance.

Develop reusable platform tooling and infrastructure automation so internal teams can provision and use compute without repeatedly solving the same underlying problems.

Improve capacity management, autoscaling and resource allocation, including namespace management and quotas, with attention to efficient use of compute.

Help make workloads and platform capabilities easier to manage across clusters and cloud environments, including the foundations for multi-cluster mobility.

Make security, patching and compliance part of the platform’s day-to-day operation, rather than separate manual processes.

Work with teams using the platform to understand friction, prioritize improvements and build components that serve multiple use cases rather than one application alone.

Own the reliability of what you build, including incident response through an on-call rotation, and turn operational lessons into lasting improvements.

Take ownership of a defined area, make progress independently and seek the context or help needed to unblock work. Adapt when operational needs or company priorities change.

You might thrive in this role if you

Have at least five years of experience building core infrastructure and can explain the systems you personally designed, implemented and operated.

Bring strong software engineering skills and experience building shared cloud or cluster platforms. Kubernetes experience is especially relevant; strong adjacent platform-building experience can also be valuable.

Have worked on problems such as cluster lifecycle, infrastructure automation, capacity, autoscaling, tenancy or safe production changes, with a clear understanding of the tradeoffs involved.

Can move between architecture and implementation, and are comfortable owning a system after it reaches production.

Build tools with their users in mind, balancing useful abstractions with the ability to diagnose and operate the underlying system.

Work effectively with ambiguity, changing priorities and incomplete information. You make progress, communicate tradeoffs and know when to ask for help.