Thomas Dohmke spent nearly four years running GitHub and helped turn Copilot into the most widely used AI coding tool globally. Now he’s starting over—with a very different thesis about where software development is headed.
His new company, Entire, emerged from stealth today with a $60 million seed round. Silicon Valley firm Felicis led the deal and called it the largest seed investment ever for a developer tools startup. The list of backers reads like a who’s who of tech: Madrona, Microsoft’s M12 venture arm, Basis Set Ventures, Yahoo co-founder Jerry Yang, Y Combinator CEO Garry Tan, and Datadog CEO Olivier Pomel. Developer community figures Gergely Orosz and Theo Browne also participated. The round values Entire at $300 million.
The Core Idea
Dohmke’s argument is straightforward: the tools developers use today were designed for humans writing code. But that’s not how software gets built anymore. Teams are increasingly managing fleets of AI coding agents that generate large volumes of code. The existing toolchain—version control, code review, and CI/CD—wasn’t built for that workflow.
Entire wants to rethink the software development lifecycle for a world where AI agents are the primary code producers, not humans.
“Just like when automotive companies replaced the traditional, craft-based production system with the moving assembly line, we must now reimagine the software development lifecycle for a world where machines are the primary producers of code,” Dohmke said. “This is the purpose of Entire: to build the world’s next developer platform, where agents and humans collaborate, learn, and ship together.”
First Product: Checkpoints
Entire’s first open-source release is a command-line tool called Checkpoints. It addresses a real problem: when an AI agent writes code, the reasoning behind the changes is often lost. Checkpoints records the instructions and reasoning behind AI-generated code and stores that context alongside the code itself. This gives teams visibility into how and why changes were made—something that matters a lot when you need to review, debug, or audit AI-written software.
Checkpoints launches with support for Anthropic’s Claude Code and Google’s Gemini CLI. Support for other popular agents is planned.
This is a practical starting point. Anyone who has tried to review AI-generated pull requests knows the pain of staring at code with no clear explanation of the intent behind it. Checkpoints address that gap directly.
Why This Matters for DevOps Teams
The shift Dohmke is describing has real implications for how DevOps teams operate. When AI agents write most of the code, the review and governance layer becomes critical. You need to know what prompted a change, what the agent’s reasoning was, and whether the output meets your standards.
Current DevOps pipelines assume a human wrote the code and understands the context. That assumption breaks down fast when you have multiple agents generating code across a codebase.
“We are approaching a tipping point where the code-centric IDE paradigm stops scaling. As AI agents become the primary producers of code, developer experience shifts from editing files to directing, supervising, and validating automated work. Thomas Dohmke is right to call for a factory reset because the existing toolchain was designed around human authorship, not machine-scale generation,” according to Mitch Ashley, VP and Practice Lead, Software Lifecycle Engineering, The Futurum Group.
“What Entire is signaling is a move toward an agent-optimized development experience where intent, reasoning, and governance travel with the code. That reframes DevOps from managing commits to managing agent behavior, accountability, and evidence. Tools like Checkpoints point to where value concentrates next: making agent-driven development observable, reviewable, and governable at scale.”
A Crowded Market
Entire enters a competitive space. Google, OpenAI, Anthropic, Microsoft, Cursor, and others all offer AI coding tools and platforms. However, most of those focus on the code-generation side. Entire is betting that the bigger opportunity is in what happens after the code gets written—the review, governance, and collaboration layer.
That’s a different bet, and it makes sense given Dohmke’s experience at GitHub. He knows the developer workflow better than most, and he understands that tools succeed when they fit naturally into how teams already work.
The Team
Dohmke moved to the U.S. after selling his startup HockeyApp to Microsoft in 2015. He became GitHub CEO in 2021 and led the company for nearly four years before leaving last August to build Entire.
The company has 15 employees, all remote, with team members who previously built developer tools at GitHub and Atlassian. They plan to expand as they work toward a broader platform launch later this year.
Entire is based in Bellevue, Washington, which puts it squarely in the middle of the Pacific Northwest’s developer tools ecosystem.
What to Watch
The real test for Entire will be whether they can build a platform that teams actually adopt. Open-sourcing Checkpoints is a smart first move—it gets the product into developers’ hands without a sales cycle. But the long-term play is clearly a broader platform, and that’s where the $60 million comes in.
If Dohmke is right that the software development lifecycle needs to be rebuilt for AI agents, Entire has the funding, team, and credibility to make a serious run at it. If he’s wrong, $300 million is a lot of valuation for a command-line tool.
Either way, this is a company worth watching.
