News

Former GitHub CEO Thomas Dohmke Launches $300M AI Coding Startup Entire

· Axios

Thomas Dohmke, former CEO of GitHub, has emerged from his August departure with a bold new venture that could reshape how developers work in an AI-dominated coding landscape. His startup, Entire, has secured $60 million in seed funding at a $300 million valuation, positioning itself as a platform designed for the reality where AI agents write most code. The funding round, led by Felicis with participation from Madrona, Microsoft's M12, and Basis Set, signals strong investor confidence in Dohmke's vision. Notable individual investors include developer advocate Gergely Orosz, former Yahoo CEO Jerry Yang, and Y Combinator's Garry Tan, bringing both capital and industry credibility to the venture. Entire's first product, Checkpoints, addresses a critical gap in the current AI coding ecosystem. The tool captures the context behind AI-written code, including the prompts, reasoning, and decisions made by agents. This addresses a growing challenge as AI-generated code becomes more prevalent but lacks the contextual understanding that human developers traditionally provide. Dohmke's strategy is notably collaborative rather than competitive. Unlike many AI coding startups that build their own models or agents, Entire positions itself as an integrative platform. "We are not training models or building agents, we are integrating with them," Dohmke explained. The platform will be open-source and designed to work with existing tools like Claude Code, Cursor, and Codex, offering developers a unified home for their AI-assisted development workflows. The timing reflects a broader industry shift. While automated coding has been one of generative AI's earliest success stories, the market has become increasingly crowded with startups competing alongside tech giants like OpenAI, Anthropic, Google, and Microsoft. Dohmke's approach of building infrastructure rather than competing directly with these players could prove strategically sound. Dohmke's vision extends beyond current limitations. He predicts that developers will soon stop reviewing code line-by-line as AI agents become more productive than humans can effectively oversee. "We have to rethink the entire system of software production from the ground up," he stated, advocating for "an agent assembly line for software development." This perspective represents a fundamental shift in software development philosophy. While current tools still assume human oversight and control, Entire is preparing for a future where AI agents handle the majority of coding tasks, requiring new systems for management, context preservation, and quality assurance. The startup's success will depend on whether Dohmke's prediction materializes and how quickly the industry adopts agent-centric development workflows. With his GitHub experience and strong financial backing, Entire is well-positioned to influence this transition, potentially establishing new standards for AI-assisted software development infrastructure.

The AI coding revolution has found its next major player as Thomas Dohmke, former GitHub CEO, announces the launch of Entire, a startup that has already secured $60 million in seed funding at a $300 million valuation. The venture represents a significant bet on a future where artificial intelligence agents handle the majority of software development tasks.

Dohmke's departure from Microsoft-owned GitHub in August, citing entrepreneurial ambitions, now makes strategic sense. His new company addresses what he sees as a fundamental mismatch between current development tools and the emerging reality of AI-dominated coding workflows. "Entire is a new platform for developers in this era where agents write most of the code," Dohmke explained to Axios.

The funding round demonstrates strong investor confidence in this vision. Felicis led the investment, with participation from established firms including Madrona, Microsoft's M12, and Basis Set. The individual investor list reads like a who's who of tech leadership, including developer advocate Gergely Orosz, former Yahoo CEO Jerry Yang, and Y Combinator's Garry Tan. This combination of institutional and individual backing provides both capital and industry credibility.

Entire's inaugural product, Checkpoints, tackles a critical challenge in AI-assisted development. The tool captures and preserves the context behind AI-written code, including the original prompts, reasoning processes, and decision-making logic used by AI agents. This addresses a growing problem as AI-generated code becomes more prevalent but often lacks the contextual understanding that human developers traditionally maintain.

The company's strategic positioning is notably collaborative rather than competitive. While the AI coding market has become increasingly crowded with startups building their own models and agents alongside tech giants like OpenAI, Anthropic, Google, and Microsoft, Entire takes a different approach. "We are not training models or building agents, we are integrating with them," Dohmke stated.

This integration strategy extends to working with existing popular tools. Dohmke specifically mentioned compatibility with Claude Code, Cursor, and Codex, positioning Entire as a platform that enhances rather than replaces current developer workflows. "Our platform will be open-source, independent, and scalable for every developer and agent to host their code and agent context," he explained.

The open-source approach could prove strategically advantageous. By avoiding direct competition with established AI coding tools and instead providing infrastructure that benefits all players, Entire positions itself as a neutral platform that could become essential to the ecosystem.

Dohmke's vision extends well beyond current development practices. He predicts a fundamental shift where developers will no longer review code line-by-line, as AI agents will produce far more code than humans can effectively oversee. "Soon, developers won't look at the code anymore, as agents will write way more than humans can review," he told Axios.

This prediction implies a complete reimagining of software development workflows. Traditional practices assume human oversight and control at every stage, but Dohmke argues for "an agent assembly line for software development" that operates with minimal human intervention while maintaining quality and context.

The market timing appears strategic. Automated coding has emerged as one of generative AI's most successful applications, with tools like GitHub Copilot, Claude Code, and Cursor gaining significant developer adoption. However, as these tools become more sophisticated and autonomous, the need for better management and context preservation systems becomes more apparent.

Entire's success will largely depend on whether the industry evolves as Dohmke predicts. If AI agents do indeed become the primary code authors, infrastructure for managing, contextualizing, and coordinating their work will become essential. Dohmke's GitHub experience provides valuable insight into developer needs and workflows, potentially giving Entire an advantage in building tools that developers will actually adopt.

The startup enters a competitive landscape but with a differentiated approach. Rather than building another AI coding assistant, Entire focuses on the infrastructure needed to support an AI-first development environment. This positioning could prove prescient if the industry continues its rapid evolution toward agent-driven development.

Referenced links