Built around security from day one. Not bolted on afterwards.
Your knowledge lives in its own environment, in the EU, under your control. This is what we do and what we do not claim.
Several systems, one environment of your own
It is often about more than one system: email, calendar, CRM, bookkeeping, documents. We connect them securely to your own environment. That is where your knowledge lives, in the EU. To answer, we use an AI model from Anthropic by default, and a person decides.
Schematic. Not a depiction of the technical set-up.
What we do, in plain words
These are our claims. No more, no less.
Your AI and knowledge base run shielded, separate from other clients. We keep project documents on our own isolated server.
Our servers are in Germany and Finland. For the AI we work with Anthropic by default. Processing may take place in the US, based on standard contractual clauses. Prefer to keep everything within the EU? We will discuss it together.
A data processing agreement, as little data as possible, and transparency about every party that works with your data.
Your data stays yours, you see what your AI knows and a person makes the important decisions.
A detailed incident protocol. If something goes wrong with your data, you hear about it quickly: within 4 hours for a serious incident, otherwise within 24 hours. That leaves you plenty of time for any report to the Dutch Data Protection Authority.
We back up our own server every night, with an encrypted copy at a second party in another location. We test that restoring really works.
The EU AI Act has a strict category for AI that helps make decisions about people, such as about work or dismissal. For our systems in employment and dismissal law, we assessed whether they fall under it. It is a borderline case, so as a precaution we treat them as high-risk and work towards the strictest requirements. We do the same with every choice about security and privacy: when in doubt, we choose the safest option.
We do not claim quality marks we do not have.
The AI prepares. You approve.
A draft reply, a draft newsletter, a proposal: your AI prepares it, a person reviews it and decides. And your knowledge is not used for anything else.
Which AI model, and where does your data go?
If you wish, we choose together which AI model we use and where your data goes.
Questions about security? Ask away, the critical ones too.
We would rather have an honest conversation than a glossy brochure. Book an intro call or email us, we reply within two working days.
What could your AI take over tomorrow?
Enter your website and see it in half a minute, with what it delivers.
Prefer to talk first? Book an intro call.