Pomegra
e2e, formerly TesterArmy, raised an $8M seed led by Credo Ventures to build a verification layer for code written by AI coding agents like Claude Code.
Key Takeaways
- e2e, formerly TesterArmy, announced an $8M seed round on November 3, 2026, bringing total funding to about $9.2M.
- Credo Ventures led, with KAYA, Point Nine, Baobab Ventures and Script Capital. Valuation is undisclosed.
- The open-source framework passed 1M npm downloads and 12,000 GitHub stars within weeks of its October launch.
Lead
e2e, the Warsaw and San Francisco startup previously called TesterArmy, announced an $8 million seed round on November 3, 2026. The company is building a verification layer for software written by AI coding agents. Credo Ventures led the round, with KAYA, Point Nine, Baobab Ventures and Script Capital participating. Angels include Snowflake co-founder Marcin Żukowski and Uber co-founder Oscar Salazar.
The company did not disclose a valuation. The seed arrives roughly three months after a $1.2M pre-seed backed by Y Combinator, which makes the new round more than six times the size of the first. Total raised now stands at about $9.2M.
What Does e2e Actually Build?
e2e is an open-source TypeScript framework, released under the Apache-2.0 license, for testing web, iOS and Android applications with AI agents. It connects to Claude Code, Cursor and Codex through an agent skill and an MCP server. A coding agent can explore a running app, write tests against interface locators and read failure reports without a human in the loop.
The framework caches actions to cut repeated model calls. It also runs against local models and existing ChatGPT or GitHub Copilot subscriptions, which keeps the cost of running it low for individual developers. It displaces hand-written suites built on tools like Playwright and Cypress, and the newer wave of hosted AI testing services.
The rename reflects a wider plan. Testing is the first piece of what the founders describe as a verification stack for agent-generated work.
Why Does Verification Matter Now?
Verification matters because agents now produce code faster than teams can review it. "Agents write more code than people, and faster than anyone can review it," CEO Szymon Rybczak said at the announcement. The bottleneck has moved from writing software to confirming that it works.
That argument has an obvious commercial logic. A coding agent that can also prove its changes work is more useful than one that only emits diffs. The skeptical reading is that the model vendors and editor makers have every incentive to build this capability themselves. A standalone verification layer has to stay ahead of the tools it plugs into.
What Traction Does the Company Have?
e2e reports more than 1M npm downloads and more than 12,000 GitHub stars since the framework's pre-1.0 release on October 1, 2026. Daily downloads have run near 100,000, and about 60 external contributors have submitted code. The company says users have run millions of tests through it.
Those are adoption figures for a free tool, and the company has not disclosed revenue or paying customers. Download counts also tend to inflate when agents install packages automatically, so they say less about human users than they once did. The metric that will matter is conversion to paid infrastructure.
How Will e2e Make Money?
e2e plans to charge for hosted infrastructure and enterprise services, not for the framework. A cloud platform for running tests at scale is scheduled for 2027. The company is also building an enterprise offering with forward-deployed engineers who handle migration and implementation.
Legacy TesterArmy customers must migrate to the new product by January 15, 2027. The seed money goes toward hiring engineers in both cities, growing the open-source team and building the cloud infrastructure.
Who Are the Founders?
Rybczak and co-founder Oskar Kwaśniewski started the company in early 2026. Both are long-time open-source contributors from the React Native community. Rybczak became the project's youngest code contributor at 16, and Kwaśniewski has landed more than 100 changes in its core.
Prior backers include Guillermo Rauch, founder of Vercel, and Walden Yan, co-founder of Cognition. Investors in the Czech and Polish ecosystems have called it one of the more closely watched young teams in the region.
What Comes Next?
The 2027 cloud launch is the test of the model. Open-source adoption has been fast, but the move to revenue depends on whether enterprises pay for managed test execution when the framework itself is free. The company also needs to keep pace as coding agents add their own built-in checks.
A second risk is that the seed was priced on momentum, not income. Three months from pre-seed to a $8M round signals investor appetite for the verification category more than proof of a business.
Outlook
e2e has raised $8M to turn a fast-growing open-source testing framework into infrastructure for verifying agent-written code. Adoption numbers are strong, while revenue, valuation and enterprise demand remain unproven. The next 12 months, including the 2027 cloud launch and the January 15 customer migration deadline, will show whether the verification layer becomes a product or stays a feature.
