- Client
- VisitAarhus
- Year
- 2026
- Industry
- Tourism
- What we did
- Research
- Workshops
- Observation
- AI prototypes
- Validation testing
A guest at the front desk wants dinner, a rainy-day plan or the bus back to the campsite, and the receptionist has seconds to answer. VisitAarhus asked us to find out whether an AI assistant could help in that moment, and what the right one would look like.
This project was research first. We facilitated workshops with hotel staff and watched how receptions answer guests day to day, then turned what we learned into working prototypes that each tested a different answer to the same question. To test them for real, we built a chatbot ecosystem around them: logins for each hotel, sharing to guests, feedback on every answer and an admin that traces each report back to the conversation behind it. Receptionists used it at their own desks, so every idea was validated in daily work, not in a meeting room.
The strongest direction came out clear: a system that merges a chatbot with more traditional tools for hotel staff. It answers from VisitAarhus’ own texts, GuideDanmark’s location data and Kultunaut’s events, checks the facts that change against live Google data, and lets the receptionist send a route to the guest with a note in Danish, German or English. The result is a roadmap VisitAarhus can build a real product on.