This article is a part of Poland Unpacked. Weekly intelligence for decision-makers
A few months after its previous funding round, TesterArmy has raised another USD 8m. But the startup is changing everything - its name, product and approach to business - its founder Szymon Rybczak tells XYZ.
Less than two months ago, we reported on the success of Polish startup TesterArmy, founded by Szymon Rybczak and Oskar Kwaśniewski. The company had raised USD 1.2m from investors in a pre-seed round. It has now secured a further, much larger investment.
“There was no pressure to raise another round. It was our decision. We saw the momentum we had, and the terms we were offered were very good,” Rybczak tells XYZ.
The company has just closed a somewhat unexpected seed round. Alongside the new funding, it is changing both its direction and its name. TesterArmy has become e2e and has now raised a total of USD 9.2m, or approximately PLN 36.5m, from a group of well-known investors.
TesterArmy’s new round
Three venture capital funds participated in the seed round: Credo Ventures, KAYA and Point Nine. They were joined by business angels Marcin Żukowski, co-founder of Snowflake; Matt Swulinski, associated with ViktorAI and previously with Wispr Flow; and Jakub Pawelski, founder of LiveKid.
“There is no hiding the fact that USD 8m lets you do a lot more, a lot faster, than a million. Today, we are no longer competing only with a few dozen startups developing AI agents for testing, but with the biggest technologies, such as Microsoft’s Playwright. We need a bigger team for that,” Rybczak says.
The e2e co-founder does not disclose the company’s valuation after the deal, but says it has risen since the previous funding round. Earlier backers of TesterArmy included Y Combinator, AIP Seed, Eight Capital and Script Capital. The investor group has also been joined by Guillermo Rauch, founder and CEO of Vercel; Fryderyk Wiatrowski, CEO and co-founder of Viktor; Charlie Cheever, head of Expo and co-founder of Quora; and Mike Grabowski, co-founder of Callstack. Bartek Pucek and Piotr and Tomasz Karwatka have also invested in the company.
A round in a week: why did the startup raise more money?
“Since returning to Poland from Y Combinator in June, we had been asking ourselves how to stand out even more. Competition was so intense that I saw a new company in this space almost every day. We were looking for something that would make someone look at our product and say, ‘They have something nobody else does.’ After hundreds of conversations with customers, including people responsible for products and engineering teams, we used that knowledge to create e2e. It still solves the same problem, just in a more mature form,” Rybczak explains.
The e2e framework debuted on October 1. According to the company, posts about it generated more than 4 million views in total, the project approached 13,000 stars on GitHub, and the software package was downloaded more than a million times. Within a month, it had been used to run 10 million tests.
“I had never seen numbers like that, even though we had worked on various open-source projects before. When we saw that level of interest, investors started contacting us. Both sides agreed it was the right time for a round. The goal was to unlock the ability to build faster. We closed it in a week. We told investors: we can keep talking, but we have one week and we are moving forward with whatever we raise in that time,” Rybczak says.
Some funds already knew the company and had been following its progress. That helped the founders close discussions within their deadline.
“Every day spent talking to investors is a day when we are not building the product. Y Combinator strongly emphasizes that if you are raising money, you need to focus fully on that. And when you are building a product, you should stay as far away from fundraising as possible, because it is distracting. We were in a strong enough position to close the round in a week and return to working on the technology the next day,” the company’s co-founder explains.
Expert's perspective
How does open source affect a business?
I expect agents to replace a substantial part of manual work and some of today’s tools. Independent verification will still be needed to ensure an agent does not mistake its own error for a “feature.” But I would go further: eventually, agents will not only detect bugs but also fix them and check whether the fixes work. This opens the door to making money from infrastructure that allows the entire process to be carried out safely.
Open source gives a startup a real chance to compete with the giants: it makes it easier to attract users, build trust and spread the technology. At today’s pace of software development, the code itself is increasingly difficult to sustain as a lasting advantage. Red Hat showed that a huge business can be built around open technology by offering reliability, support and predictable performance. I have always been particularly interested in companies where software is a means of delivering a broader service. In this case, the customer is buying confidence that their software works and that detected bugs are fixed quickly. The code can be made available for free, while the company earns money from the infrastructure and from taking responsibility for the quality of the entire service. Success will depend on whether the startup can turn its popularity among developers into value that businesses are willing to pay for.
The end of TesterArmy: the company goes all in on e2e
The name change is more than a rebrand. The existing TesterArmy product is to be phased out, even though it already has paying customers. The company is focusing on a new solution - e2e - which it is building from scratch and which will become TesterArmy’s successor.
“At this stage of the company’s development, we have to make big decisions. We worked on TesterArmy for several months and already had more than 70 paying customers, yet we are migrating the entire product. From a technical perspective, we can see that what we have created with e2e is much better. TesterArmy had to exist for us to reach the next version. We had to understand what works, what does not and what could be done differently,” Rybczak says.
The biggest change is the release of the source code. e2e is being developed as an open-source project, available free of charge to developers and companies. Users can download the framework, connect it to an AI model of their choice and, if necessary, change the AI agent or the environment in which they run tests. They do not have to rely exclusively on infrastructure provided by the startup.
The framework is also designed to reduce the number of calls to the AI model. On the first run, an agent determines how to perform a task, and e2e records its actions. During subsequent tests, the framework replays that sequence without involving the model again. If a change to the application causes a step to stop working, the agent determines how to perform it again. The approach is intended to combine AI’s flexibility with a repeatable testing process.
“We want e2e to become a standard and a verification step for every piece of code created. We did not have that opportunity with the previous product. Throughout software history, the most important projects have very often been open source. Agents also prefer open-source tools that can be used for free. We designed e2e to be ready to work with agents, because they are increasingly involved in choosing tools,” the startup’s co-founder says.
Rybczak says a significant part of the growing interest in the project is already coming from AI agents that choose tools independently.
Expert's perspective
The wave is coming
Free code, paid infrastructure
The startup plans to make money from paid infrastructure for running tests and from solutions developed for large enterprises.
“Revenue on a larger scale should arrive next year, although some is already coming in. We did not expect to start talking to large enterprises so quickly. Open source makes the decision easier for them because it reduces the fear that a provider might disappear and take the technology with it. The code is on GitHub, and a customer can create its own version or build something on top of it at any time. That is especially important for the largest companies,” Rybczak says.
The first paid offering is expected to be the e2e cloud. The startup plans to provide browsers, simulators and physical iOS and Android devices on which teams will be able to run tests. A second line of business will consist of deployments tailored to the infrastructure of large enterprises. The knowledge gained from those projects is then expected to support development of the open framework.
“e2e will always be available to use for free. Our cloud will be optional. If someone has their own infrastructure or wants to use another provider’s solution, they will be able to do that. Our service will be one of many options. In addition, the largest enterprises sometimes need a tailored version of the product. Those are the two paths to monetization we see,” the startup’s founder explains.
Competitors have also expressed interest in the framework. The company has already begun working on integrations with some of them.
“If e2e becomes a standard in a world where agents write code, we will have achieved our goal. Someone else can make money from it too. We will build a business on our own cloud, but we do not want to close the entire market off for ourselves. Developers quickly recognize projects that are open in name only and force users to rely on one provider at a larger scale,” Rybczak says.
Expert's perspective
How AI agents will change the market
But remember that not every open-source project is a good business. It is the cheapest distribution channel because developers bring the product into companies themselves. Red Hat, acquired by IBM for around USD 34bn, is a model, while MongoDB, Elastic and GitLab show that multibillion-dollar companies can be built. When assessing such a company, I look at three things. The first is a clear boundary between what is free and what companies pay for. The second is the risk that a hyperscaler commercializes the product faster than the startup itself. The third is the conversion of adoption into revenue, because GitHub stars are not a business model.
From eight people to fifteen: the company prepares to build its cloud
The most immediate effect of the new funding round will be an expansion of the team from its current eight people to around fifteen over the next few months.
“We are primarily hiring engineers. We want to attract developers who have worked on important technology projects, particularly open-source ones. It is a specific kind of work, different from creating code solely for one company. At the same time, we are building a team to create our cloud for running tests. That is a new area for us. We also need people who will work directly on deployments for large customers,” the e2e co-founder says.
Companies on the Fortune 500 list are said to be interested in the Polish tool, although Rybczak does not want to disclose their names. Discussions with potential customers are also intended to help the startup better understand the needs of large organizations. As it develops its paid offering, e2e plans to build a sales team gradually.
Recruitment will take place around the world. The company operates primarily remotely and relies on asynchronous communication. It has an office in Warsaw and plans to open another in San Francisco. It also wants to attract Polish developers.
Expert's perspective
What is happening in AI?
Running open-source projects is harder today than ever. “Free” is no longer an advantage - technology giants also make their solutions available at no cost. Code has become cheaper: an AI agent can recreate someone else’s project in a few days, so the code itself is no longer a barrier to entry or a source of value on its own. It is still possible to win in the market, but not with code alone. The winner will be the one that builds a community before the giants notice the niche.
More investment ahead
At the same time, the company plans to develop features for different platforms, from web browsers to mobile devices and desktop applications. Its plans include expanding the ability to test gestures, animations and application performance - for example, how smoothly an interface runs. The founders want the framework to support as many important testing scenarios as possible.
“We want everything that can be done in a browser or on a phone to be testable too. Every gesture, every animation. We are also looking at performance - for example, whether an animation runs smoothly at 60 frames per second after a button is clicked. We want to create the most powerful framework possible for testing any software. There is a huge amount of work to do, and thousands of teams are already using the tool, so every feature and how it is used matter,” Rybczak says.
By the end of the year, however, the company’s main goal is to increase awareness of its technology among AI agents.
“We would like e2e to be the first recommendation when someone asks an AI agent on January 1 what the best way to test an application in 2027 is. We are already seeing a few such recommendations. But we want to reach a point where new models know our technology and agents naturally turn to it. That is the direction in which we want to develop the product,” concludes the co-founder of TesterArmy, now e2e.
Key Takeaways
- Polish startup TesterArmy, founded by Szymon Rybczak and Oskar Kwaśniewski, has raised another USD 8m just two months after its previous round. The company has now raised USD 9.2m in total, or approximately PLN 36.5m. The seed round included Credo Ventures, KAYA and Point Nine, as well as prominent business angels. Investor discussions lasted just one week. The founders decided to raise additional funding after strong interest in their new solution, which was used to run 10 million tests in the month after its launch.
- Alongside the new funding, the startup is radically changing its strategy. Its existing TesterArmy product, which had more than 70 paying customers, is to be phased out. The company is focusing on e2e, an open-source framework that uses AI agents to automate software testing. The technology is intended to reduce the cost of using AI models and allow developers to choose their own infrastructure. The framework itself will remain free. The company plans to earn money from cloud services and solutions for large enterprises while keeping the core technology open.
- The new funding is intended to help e2e accelerate development of its technology and expand its team from eight to around fifteen people. The startup plans to build its own infrastructure for application testing, develop features for mobile devices and computers, and expand into the US market. The company is already in talks with Fortune 500 businesses. Its long-term ambition is to create a software-testing standard for a world in which AI agents write an increasing share of code. The founders want their tool to become a natural choice for both developers and the agents themselves.