At Madrona, we believe the most important technology and market shifts require founders who don’t just understand the change — they’ve lived at its center and driven the change. This exactly describes Thomas Dohmke, the former CEO of GitHub, who grew Copilot into a multi-billion dollar business and scaled the platform to over 180 million developers.
Today, we’re thrilled to announce our investment in Thomas’s new company, Entire, which is building the developer platform for the AI era where agents and humans collaborate, learn, and ship together. Entire has raised a landmark $60 million seed round, the largest seed investment in developer tools history, led by Felicis, with a significant check from Madrona, and participation from Microsoft’s M12, Basis Set, and other influential investors.
A Founder We’ve Known for Over a Decade
Our investment in Entire is fundamentally a bet on Thomas and his vision. Madrona’s relationship with Thomas goes back over a decade, from Soma Somasegar leading the acquisition of Thomas’s previous startup, HockeyApp, for Microsoft’s Developer Division, to working alongside him in DevDiv, to watching Nat Friedman handpick him to lead GitHub. Tim has served on the board of Spice AI with Thomas since its formation and has been consistently impressed by his rare balance of strategic breadth, bias for action, and no-nonsense ability to get things done. He ran a massive organization inside one of the world’s largest corporations without letting it lose its startup ethos.
Thomas also has a rather unique set of insights into what made GitHub great and what was holding it back. He knows every major player in the code-gen landscape intimately, their gross margins, business models, technical approaches, and partnership strategies. And he remains deeply connected to the open-source community that was essential to GitHub’s rise. When Thomas came to us with the idea for Entire, he and his co-founders arrived with a crystal-clear vision on Day 1. It was immediately obvious: this team knows exactly what they are doing. Madrona was the first investor to commit.
The Problem: Software Production’s Manual System Is Cracking
For decades, software development had a certain flow. You had an idea, opened your editor, changed a few files, pushed a commit, and asked for a code review. Files, folders, diffs, pull requests — that was our little workshop. Line by line, human to human, software production moved at the speed of thought.
That order no longer exists. We are living through an agent boom, and massive volumes of code are being generated faster than any human can reasonably understand. Developers now orchestrate fleets of agents across multiple terminal windows. Specs drive code generation. Agents interoperate in parallel, generating hundreds of variants simultaneously.
The truth is: our manual system of software production — from issues, to Git repositories, to pull requests, to deployment — was never designed for the era of AI. GitHub and Git were architected for human-to-human collaboration in the cloud era. Version control preserves change, but it does not preserve meaning. The intent, constraints, and reasoning that explain how decisions were made simply don’t exist in today’s tools.
GitHub is an incredibly important platform, but it is quickly becoming a legacy platform. Consider the pull request: how does reading code line by line scale when agents are working around the clock producing hundreds of variants? It doesn’t. The answer isn’t to bolt something on top. It’s to reinvent the paradigm. We know GitHub has massive network effects with hundreds of millions of repositories. But we deeply believe in Entire’s vision that now is the moment to reimagine the entire software development lifecycle for a world where machines are the primary producers of code and that AI agents and humans need to interact efficiently. And that this creates the need and opportunity to build a new platform for the era of AI agents.
The Solution: A New Assembly Line for Agents
Entire is reimagining the developer platform for the AI era from the ground up, built around three core components:
Version control for agents with a Git-compatible database that goes beyond storing source code to become an intelligent version control system of record, capturing prompts, models, datasets, evals, intent, and agent context in a single place. Agents move fast, humans retain governance, and every decision traces back to intent, not just code commits.
A universal semantic reasoning layer that unlocks agent-to-agent collaboration with persistent, shared, pluggable memory tied directly to version control history, so fleets of agents and groups of humans can reason, coordinate, and build in parallel without overwriting or losing understanding.
An AI-native software development lifecycle with a new UI where developers can review, approve, and deploy hundreds of changes per day without becoming the bottleneck, unlocking human creativity by making software production dramatically cheaper and faster.
The platform is open, scalable, and independent, regardless of which model or agent you use, and will be open source at its core, unlike GitHub. Thomas also brings both open-source sensibility and deep enterprise relationships. He plans to serve both audiences from day one, working with enterprise design partners while building the single-player attraction for open-source developers globally.
Critically, Entire is compatible with GitHub from the start. Developers won’t need to abandon existing repositories to begin using Entire’s AI-native capabilities. This provides a natural on-ramp while the team builds toward its massive ambition — not only to supersede GitHub, but to superset it, capturing the broader set of developer tools and services currently fragmented across the ecosystem.
Why We’re Excited
Madrona has been studying the shifts in AI code-gen and evaluating new companies across the software development lifecycle — the market, the opportunities, and the challenges. We are eyes wide open about both the size of the prize and the difficulty of grabbing it. Code-gen adoption is arguably the fastest-growing use case in generative AI, creating powerful tailwinds for a platform purpose-built for this new reality. GitHub demonstrated what’s possible: a developer platform that became essential infrastructure for the entire software industry. The next platform has the potential to be even larger.
Thomas Dohmke is a special founder with differentiated, perhaps unique, insight into a very large and well-timed opportunity. His strengths as a startup founder — action-orientation to the point of impatience, talent magnet, deep empathy for the customers and users — are exactly what’s needed to reimagine a market dominated by entrenched incumbents. He has assembled a world-class team of more than 15 developers, veterans of GitHub and Atlassian, building fully distributed across the U.S., Europe, and Australia. Just builders, moving fast.
Soma and Tim will be working closely with the company, with the help of the entire Madrona team, and Tim will be joining as a board observer. If you’re a developer who wants to be part of building the future of software development, visit entire.io and join the movement. The new era of software development begins now.