How we work
Seven rules for what we take on, what we refuse, and how we build. They are here so you can hold us to them.
Who runs it is your choice, not our revenue model.
We are happy to host it for you, but source code, deploy scripts and a runbook are included in the first release regardless.
A choice you cannot explain, you cannot reverse.
Architectural decisions are recorded as RFCs, with the trade-off and the measurement included.
Software is done when it survives a year without us.
At the start we write down when we are redundant and what needs to happen before that.
Your data does not leave the EU without your explicit instruction.
We host in Utrecht. No US cloud service sits in the critical path. That is a hard requirement for part of our public procurement work.
Everything we run ourselves is public.
If you are our client and want to leave tomorrow, you can host the whole thing yourself. The source code is on GitHub, the deploy instructions are in the repository.
We say no to work that conflicts with these principles.
That includes projects that require closed-source forks of our open-source code, integrations with US cloud services in the critical path, or deadlines that force us to skip the RFC step.
A project has an explicit end date, and we stick to it.
We write down at the start when we hand over and what that handover looks like. Not as a threat, but because software without an exit plan is a dependency that keeps getting more expensive.