Former GitHub CEO Thomas Dohmke’s new startup, Entire, launched a preview of its distributed Git network on Wednesday to relieve the mounting pressure autonomous AI coding agents place on centralized developer infrastructure. The Bellevue, Washington-based company is deploying initial network regions across the United States, Europe, and Australia to keep code repos accessible as automated systems overwhelm traditional servers.
The Bottleneck of Centralized Code Repositories
Git was originally created in 2005 by Linux founder Linus Torvalds as a decentralized version control system. However, the software industry over the last decade consolidated around centralized cloud hosting platforms like GitHub, which Microsoft acquired for $7.5 billion in 2018.
The recent explosion of AI coding assistants—including Claude Code, Cursor, Codex, and GitHub Copilot—has altered developer workflows dramatically. Rather than humans making occasional updates, autonomous software agents continually query, clone, and pull code at non-human speeds.
This surge in activity creates server congestion, causing frequent rate limits, high latency, and unexpected system outages for engineering teams worldwide. To solve this, developers need infrastructure capable of scaling alongside machine-driven interaction.
A Distributed Mirror for the Era of AI Agents
Entire’s new platform allows development teams to mirror existing GitHub repositories in a single step without transferring primary ownership. AI agents can then read and pull data from local, geographically closer network mirrors rather than hammering a single central server.
In announcing the launch, Dohmke referenced Torvalds’ classic 2007 principle: ‘If you’re not distributed, you’re not worth using.’ Dohmke emphasized that centralized Git hosting has become a fundamental constraint in the era of artificial intelligence.
In addition to repository mirroring, Entire addresses the growing challenge of tracking AI decision-making. The platform records session logs, capturing the exact prompt instructions, reasoning steps, and context behind every AI-generated line of code directly inside the repository.
Accompanying Wednesday’s launch are several specialized developer features: Entire Blame traces code lines back to the specific AI conversation that produced them, Entire Review conducts automated code reviews using that historical context, and Code and Semantic Search allows teams to query the reasoning behind past updates.
Backed by Major Venture Capital and Industry Rivals
Entire emerged from stealth in February with a $60 million seed round led by Felicis, marking the largest seed investment ever for a developer tools startup and valuing the company at $300 million. Additional investors include Madrona, Microsoft’s venture capital arm M12, Basis Set Ventures, Yahoo co-founder Jerry Yang, and Y Combinator CEO Garry Tan.
Industry analysts point out that Entire’s non-competitive stance toward existing tools strengthens its market positioning. The platform integrates seamlessly across every major AI coding agent rather than competing directly against them.
‘We think it can be the next great developer platform,’ said Tim Porter, managing director at Madrona. Porter noted that working in conjunction with tools like Cursor and Claude Code provides Entire with a unique advantage.
The participation of Microsoft’s M12 fund highlights a cooperative dynamic between traditional cloud providers and next-generation infrastructure. While GitHub remains an industry standard, venture capital partners argue that central platforms must adapt or risk becoming legacy infrastructure in an agent-dominated ecosystem.
Implications for the Software Industry
Entire’s immediate focus centers on code mirroring, but the startup plans to eventually allow developers to host new repositories natively on its distributed network. The fully remote company has scaled to more than 40 employees across nine countries as it prepares for wider adoption.
Following the preview period, Entire intends to roll out commercial and individual pricing tiers, blending seat-based and consumption-based subscriptions alongside open-source elements and a free tier. As software development transitions from human-written code to AI-orchestrated workflows, session logs and distributed networks are set to become foundational elements of modern software engineering.
