Forum Libraries Nacka. What a live voice agent actually sounds like.
A two-minute look at a Digital Venue agent that answers thousands of real questions a week at a three-branch Swedish library network — in voice, in text, in every language a visitor might walk in speaking.
What you're watching.
This is real, not a mock-up. The agent in the video is the same one anyone can talk to right now at forumbiblioteken.alex2.digitalvenue.ai. It runs on EU infrastructure, grounded in the Nacka Forum library catalogue, and answers in voice as well as text — in whichever language the visitor speaks first.
It's not a voice front-end on top of a generic chatbot. It's a single conversational layer that hears, understands, retrieves from the venue's own data, and replies in a voice the library chose.
How it ships in 15 minutes.
- State your mission. A sentence or two about who you are and what you want the agent to do.
- Add documents and endpoints. Drop in PDFs, link the catalogue, point at an opening-hours API. Anything visitors might ask about.
- Pick a voice. Use one of ours, or clone the curator's, the tour guide's, the stadium announcer's.
- Go live. A QR code on the front desk, a widget on the site, a phone number — your call.
Where it runs.
All inference and storage stay inside Europe. The Forum Libraries deployment is hosted in Stockholm and Amsterdam by default — no CLOUD Act exposure, no data leaving the building unless we say so. Larger institutions can take exactly the same agent on-premise.