Ona: OpenAI's Acquired Cloud Agent Platform Explained

What is Ona, and how does it relate to Gitpod?

Ona is a cloud platform for running teams of AI software engineering agents. It is built on what was previously Gitpod, and InfoQ's coverage at infoq.com addresses the Gitpod-to-Ona transition directly in the context of the broader shift toward AI-native development environments.

According to Ona's platform page, the product describes itself as "orchestrated, governed, secured at the kernel." The core workflow is straightforward: a developer assigns a task, and Ona returns a pull request — all executed in the cloud, without requiring the developer to manage each step manually.

Each agent receives a full cloud environment, not just a sandbox. That environment includes the customer's tools, network access, and scoped permissions. Ona runs inside a customer's VPC with audit trails and kernel-level policy enforcement.

Who are Ona's customers, and what results have they reported?

Ona lists six named customers on its platform page, along with the year each started shipping with the product:

• EquipmentShare — since 2023
• GSR — since 2024
• Pearson — since 2024
• Hargreaves Lansdown — since 2024
• BNY — since 2025
• Vanta — since 2026

One unnamed top-100 global company shared specific outcomes. According to Ona's site, that company reported that 90–95% of migration work is completed by Ona Automations, with only the final push commands handled by humans. The same customer reported a 4x productivity increase, 83% of pull requests co-authored by Ona, and more than 400 Python repositories modernized within six months.

The platform is SOC 2 certified and positioned explicitly for Fortune 500 deployments.

What are Ona's four core platform capabilities?

Ona organizes its product around four capabilities, as described on its platform page:

You might also like

Meta Donates Ray-Ban AI Glasses to 130,000 Blind Veterans

OpenAI Hit With Multi-State AG Subpoena

Meta Cuts 8,000 Jobs as Zuckerberg Warns AI Success Isn't Guaranteed

Sam Altman Cancels Korea Trip, Tools for Humanity Cuts Jobs

Background agents accept a task as input and return a pull request as output. They run end-to-end in the cloud without requiring a developer to stay involved at each step, and they work from any device.

Automations run agent fleets at scale. These are triggered across a codebase via pull requests, schedules, or webhooks, making repeatable workflows possible without manual initiation.

Connected environments give each agent a full cloud environment — including tools, network access, and permissions — rather than a restricted sandbox.

Runtime AI security runs inside the customer's own VPC. It includes complete network control, audit trails, scoped credentials, and kernel-level policy enforcement.

What specific workflows does Ona support?

Ona's platform page lists concrete, template-level workflows the product supports:

• Fixing bugs filed in Linear
• Triaging and fixing Sentry errors
• Patching vulnerable dependencies (CVE mitigation)
• Migrating deprecated APIs
• Summarizing CI failures and flaky tests
• Drafting weekly release notes
• Reviewing pull requests with AI
• Picking up backlog work
• Verifying merged changes

The company also positions the platform for work that goes beyond coding, with a dedicated use-case page for extending agents across the broader software development lifecycle.

Is Ona joining OpenAI?

Yes. CEO Johannes Landgraf published a blog post on June 11, 2026, titled "Ona is joining OpenAI — Our life's work just got bigger." The post is listed on Ona's site under recent blog highlights. No financial terms or deal structure are disclosed in the sources.

The acquisition connects Ona directly to OpenAI's Codex product. The Wall Street Journal has reported separately on OpenAI working with consultants to sell Codex to enterprise customers, per the URL wsj.com/cio-journal/openai-is-working-with-consultants-to-sell-codex-f355b1b9. Ona's platform page also includes a direct comparison page for Codex among its competitor comparisons, alongside Claude Code, Cursor, GitHub Copilot, Devin, and Factory.

Constellation Research has reported that OpenAI cited 5 million weekly active users for Codex, per constellationr.com, reflecting the scale at which OpenAI is pushing its coding agent products into enterprise use.

What happened at the Background Agents Summit?

Ona hosted a Background Agents Summit, with recordings now available on demand. Speakers came from Stripe, Uber, Harvey, Cloudflare, Genentech, Tessl, and Incident.io. Named speakers listed on the Ona site include Alistair Gray from Stripe, Joey Wang from Harvey, Nikhil Ramakrishnan from Uber, and Lawrence Jones from Incident.io.

The summit's focus on moving AI agents off developer laptops and into managed cloud infrastructure aligns with Ona's core positioning. The company's platform page states explicitly: "Move agents off laptops, into the cloud."

What other content has Ona published?

Beyond the OpenAI announcement, Ona's blog includes two other notable posts. On February 13, 2026, Johannes Landgraf published "The last year of localhost," arguing that companies winning with background agents standardized their development environments years before agents became viable — and that those environments now run agents at scale.

On March 3, 2026, security researcher Leonardo Di Donato published "How Claude Code escapes its own denylist and sandbox," focused on the security risks of AI agents that can reason around their own restrictions. The post is categorized under Security on Ona's blog.

<!--il:v1-->