An AI chat answers questions at any hour, in several languages, and helps visitors reach you. The problem is that most solutions send every message your customers write to an outside provider.
Where the messages go
With an off-the-shelf chatbot, whatever a customer types usually travels to the servers of a company in another country, where its model processes it. For a clinic, a law firm or any business handling sensitive data, that means risk and awkward questions about the GDPR.
The model can run on your side
Open language models are now good enough to run on your own server. The chat on this website uses one of them, running on our own servers. Messages don't go to third parties, and we decide what the assistant knows and what it doesn't.
The assistant only says what it knows
A good assistant doesn't make things up. Ours answers only from the text we write about our services, process and projects. When it doesn't know something, or someone asks about prices, it hands the conversation to a person from the team, who replies in the same chat.
Transparency
Visitors should know they are talking to a program. The European AI Act requires it, and trust requires it anyway. That's why the chat introduces itself in the first message, and you can ask for a person at any time.
Who it makes sense for
- Businesses that get many repeated questions: bookings, rough prices, opening hours.
- Clinics, service companies and online shops.
- Anyone who doesn't want to lose customers who write in the evening or at the weekend.
Want an assistant like this for your business? Try it in the corner of the page or write to us.
