Your data stays your data.

What we build runs in your own environment, inside the EU. We only use the data the process we automate actually needs.

Your data is encrypted in transit and at rest: TLS the moment it is sent, AES-256 wherever it is stored.

This website, too, is only ever served over an encrypted connection.

We work with the AI vendors' business APIs, whose terms state that your data is not used to train models.

We put that in writing with you as well, in the data processing agreement for every project, conversation data included.

We take over the permissions your systems already have. Whoever may not see something in the source system does not see it in the data layer either, nor in anything that runs on top of it.

We build and test on a copy, not in your live systems. Halfway through you get it in your hands, with real conversations and real data.

Whatever comes out wrong, we fix before a single customer meets it. It only goes live once you approve.

What we build runs in its own environment, on its own network, and is not open to the public internet.

The form on this site works the same way: it may write a message, but cannot read anything back, and the database can only be reached with a separate key.

Every step is recorded: who or what took which step, and when, readable afterwards.

If something goes wrong, the workflow stops and you get a notification, instead of wrong data quietly flowing through. Failure has to be visible.

An exception goes to a person, the rest keeps moving. You see every conversation and can step in at any moment; the agent then steps back.

If an agent does not know something, it says so and passes it on. We would rather build it a little too cautious than have it make things up.

On every project we set out how we handle your data in a data processing agreement under article 28 of the GDPR. The data stays yours; we are the processor.

You get the documentation, and you can leave us without starting over.