Former GitHub CEO Thomas Dohmke has launched Entire, a new developer platform focused on human–AI agent collaboration, backed by a $60 million seed round that values the company at $300 million.
The round was led by Felicis, with participation from Madrona, Microsoft’s M12, and Basis Set, along with angel investors including Gergely Orosz (The Pragmatic Engineer), Theo Browne (T3 Chat), Jerry Yang (AME Cloud Ventures), Olivier Pomel (Datadog), and Garry Tan (Y Combinator).
Entire is building an AI-native developer platform designed to make agent-generated software auditable, governable, and reproducible. Instead of treating AI output as transient chat responses, Entire stores agents work as first-class, versioned data.
At the core of the platform is a git-compatible database that records not only source code, but also prompts, agent reasoning, constraints, decisions, and execution traces alongside each commit.
The company’s first release, Checkpoints, is an open-source command-line tool that is designed to “automatically capture context from popular coding agents.” It automatically attaches a structured “checkpoint” to every agent-generated commit.
Each checkpoint captures the full session context—including transcripts, prompts, tool calls, and intermediate reasoning—and stores it directly inside Git, where it can be searched and visualised through a dedicated interface.
“This data is stored directly in Git and visualised through a dedicated UI. Launching today with support for Anthropic’s Claude Code and Google’s Gemini CLI, Entire will extend support to all leading agents in the coming months, integrating with the broader developer tools ecosystem ahead of its platform launch later this year,” the company said.
The startup aims to provide an AI-native user interface for reviewing and governing agent output at scale. The interface allows developers to inspect intent, trace decisions back to their originating context, approve or reject changes, and manage large numbers of parallel agent-driven modifications without relying solely on traditional pull requests.
Dohmke said Entire is being built to address a structural gap between existing developer workflows and the pace of agent-generated code.
“We are living through an agent boom, and now massive volumes of code are being generated faster than any human could reasonably understand,” he said. “Our manual system of software production—from issues, to git repositories, to pull requests, to deployment—was never designed for the era of AI.”
Investors described the deal as a bet on a new platform layer for software development. Felicis founder and managing partner Aydin Senkut said that bolting agents onto human-centric workflows is creating “friction, complexity, and real bottlenecks,” and that Entire is rethinking the developer platform from first principles for an agent-first world.
Madrona partners Tim Porter and Soma Somasegar said version control systems built for human-to-human collaboration do not preserve the intent and reasoning behind machine-generated code, making a new architecture necessary for governance, security, and large-scale coordination.
Entire is a remote-first company with 15 employees and plans to grow to more than 30 engineers in the coming months.
ALSO READ: OpenAI Begins Testing Ads on Free and Go Plans in the US
