Blumenhof Hotel is an 8-room boutique property in Bucovina, Romania — near the UNESCO-listed Voronet Monastery. It runs on an AI system built by Lypt Systems. Zero missed calls. Eleven languages handled automatically. No night staff hired. This is the story of how and why we built it.
Why we built it
Blumenhof had a straightforward problem: no night desk. When guests called after hours — which happens regularly, especially from German and Austrian travelers planning trips to Bucovina — the phone rang out. Voicemail doesn’t book rooms.
During busy periods, the front desk couldn’t answer every call either. Staff were serving in-person guests while the phone kept ringing. Every unanswered call was a potential booking walking away.
We didn’t need a chatbot. We didn’t need a new CRM. We needed a system — one that answered calls, understood what the guest wanted, captured the details, and told the team what to do next morning.
What the system does
The AI phone concierge answers every call to Blumenhof, day or night. Here’s what a typical overnight flow looks like:
A German-speaking guest calls at 11pm asking about room availability for next weekend. The AI answers in German, provides room options and pricing, captures the guest’s name, dates, and contact number, and confirms that the team will follow up in the morning with a booking confirmation.
At 8am, the front desk opens their handover report. It shows every call from overnight: who called, in what language, what they asked, what the AI told them, and what action is needed. The booking lead from the German guest is flagged as high-priority with all details ready for follow-up.
No voicemail. No lost lead. No manual transcription.
The three layers in practice
Agents: The voice AI is trained on Blumenhof’s specific room types, pricing, policies, amenities, directions, and local recommendations. It doesn’t sound like a generic bot reading a script — it speaks about the hotel the way the team would.
Automations: Every call triggers a workflow. Guest details are logged. Priority is assigned based on booking intent. The morning handover is compiled automatically. Follow-up reminders are created in the task system.
Intelligence: A dashboard shows call patterns — which hours are busiest, which languages appear most, which questions come up repeatedly. This data shapes both the AI’s training and the hotel’s operational decisions.
The results
Since deploying the system, Blumenhof has not missed a single after-hours call. Guests in 11 languages are handled without any multilingual staff. The front desk starts every morning with a clear handover instead of a blinking voicemail light.
The most telling result: German-speaking guests regularly book after calling at night. Before the AI, those calls went to voicemail. The bookings went to competitors.
Why this matters for your business
Blumenhof is our proof environment. Every system we build for clients uses the same three-layer architecture, the same integration approach, and the same pilot methodology that runs our own hotel.
We don’t demo in sandboxes. We demo in production — with real guests, real calls, and real revenue at stake.
If your business has calls that go unanswered, leads that go cold, or follow-up that falls through the cracks, the system that runs Blumenhof can be adapted for your stack and your outcomes.