Knowledge that disappears
A long-time employee leaves — and with them, nobody knows anymore how the regular guest likes to be addressed or how the PMS is configured. Institutional knowledge lives in heads, not in systems.
Level 1 — The foundation
Your hotel knows more than any AI. We bring the two together.
SOPs, house rules, briefings — your knowledge lives in Word documents and in the heads of employees who eventually leave. We make it AI-ready. Once. For everything that comes after.
Book a free intro call
30 minutes · No commitment · No pitch
The problem
The knowledge is there. It's just not accessible.
Voice agent without a base = expensive failure
Many hotels want an AI phone assistant. In 80% of cases it fails not because of the technology — but because no prepared knowledge base exists.
What we build
The foundation for every AI that comes after.
- Stocktaking of all documents, SOPs, FAQs and processes
- Structuring and cleanup (duplicates, contradictions, gaps)
- AI-ready preparation: chunking, tagging, embedding preparation (RAG)
- Building the technical infrastructure (vector database, API connection)
- GDPR-compliant documentation of the entire process
- Handover: your team understands what was built and how to maintain it
Possible afterwards
Investment
Start small. Keep growing.
Focused build
One defined knowledge source · Technical integration · GDPR docs
Duration: 3–6 weeks
3,000–10,000 EUR
Complete knowledge base
All sources · Scalable architecture · Team training
Duration: 6–12 weeks
10,000–25,000 EUR
Annual maintenance: 20–40% of the setup price. Optional.
The difference
“Data preparation is 30 to 50 percent of the effort in any AI project. That's the part where others fail. And the part we get right from the start.”
Most AI consultants show you how a voice agent works in the demo — then they hand over the technology and you realize: your hotel knowledge isn't even in it. We start where others stop.
7 years hotel GMAI EngineerFounder of avvicendaDACH + Italy
The 5 most common reasons
Why hotel AI projects fail — and how a knowledge base prevents it.
01
No hotel-specific knowledge
Voice agents without a knowledge base answer generically and incorrectly.
02
Knowledge stuck in employees' heads
When an experienced employee leaves, their knowledge leaves with them — without a database, permanently lost.
03
Documents nobody can find
An 80-page manual only helps if it's accessible — the database makes it retrievable.
04
Data preparation underestimated
30–50% of the effort in AI projects comes from structuring. Whoever skips it fails.
05
No plan for maintenance
A knowledge base that isn't maintained goes stale. With a defined maintenance process it stays usable.
Common questions
All clear.
The foundation is set. Everything else follows.
30 minutes · No pitch · No commitment