Skip to main content
All posts

Why we started a company after all

In 2023 we open sourced Kviklet. At that point it was a failed startup idea, and we later wrote that we were unlikely to ever build a company around it. In summer 2026 we founded one anyway. How come?

More traction than we expected​

After we open sourced the code, Kviklet got a surprising amount of traction in the community. People on Reddit and Hacker News liked the idea. The GitHub repository kept getting feature requests, and they were great ones. To this day we have not gotten around to implementing all of them.

Those requests also taught us two things. A lot of them showed that people were using Kviklet in larger organizations than we had originally thought. And they showed us that the problem itself was more widespread than we had thought.

That was good news, but it came with a catch. By then both of us had started working full-time again, so Kviklet was, and still is, something we work on in our free time. And it is hard to justify spending your weekends on a project that is very apparently an enterprise product.

So we had a project that people used and kept asking for more of, and no good way to give it the time it needed.

Asking before building​

In 2025 we ran a test. We wanted to find out whether people would be willing to pay for some features that they had asked for and that we had not built yet.

And it turns out we did catch some interest. So together with those first interested customers we ended up building a real enterprise version, with SAML, role sync, role-based review requirements and more.

To everyone who is still giving us continuous feedback: thank you. You are making Kviklet a great product for engineers to use. Keep it coming!

Founding the company​

Since then, the world around us has changed too. Coding agents are on the rise, and cyber security is becoming an ever more relevant concern. The number of published CVEs has shot up, and more and more companies are implementing careful security guidelines, often to the detriment of the engineers who have to go through them. That friction is the problem Kviklet was always meant to solve.

So in summer 2026 we finally founded a real company for Kviklet. With that step we are committing to the long-term support and development of the product.

So far this is not the rapidly expanding, venture-funded startup story that we imagined in early 2023. Nothing about the last three years looked like that. It was a failed idea, then an open source project, then a test to see if anyone would pay, then a few customers, and only now has it become a company.

But maybe that is fine. Maybe a stable, useful product that is loved by its loyal users, and that finally makes engineers, security and compliance folks be friends, is the better outcome for everyone.