Skip to content
A man in a white turban and brown robe beside a leopard wrapped in white fur

Systems

Chatbots for hotels and apartments

A chatbot for hotels and apartments answers guests in their own language at any hour, holds the booking, sends the deposit link and passes anything unusual to you. This is what we build for properties, what we have already shipped, and where a chatbot stops being the right tool.

A guest in Munich plans a summer week at one in the morning. She writes to three apartments on WhatsApp and asks the same two questions, whether it is free for those dates and whether there is parking. The owner who answers first usually gets the booking. At that hour, the owner is asleep.

Three moments a guest writes

Guests contact a property before they book, during the stay and after they leave. Each moment needs a different tool, and only the first one is mainly a chat.

01Before the bookingDates, parking, pets, the transfer from the airport. The system answers from what you gave it, in the guest’s language, holds the booking and sends a deposit link.
02During the stayTowels, breakfast in the room, a late checkout. These fit a menu better than a conversation, which is why we built iButler, covered below.
03After checkoutThe system asks for a review at checkout and drafts replies to the reviews that come in, for you to approve before anything is posted.

What a chatbot for hotels does in one conversation

LVila LazurAnswered by the system
  1. 01Answer
  2. 02Hold
  3. 03Deposit
  4. 04Review
WhatsApp · 01:40Hello, do you have a room for two from 12 to 15 July? And is there parking?Hello Anna. A double room with a balcony is free from 12 to 15 July, and parking in the courtyard is included. Shall I hold it for you?Yes pleaseIt is held for you. The booking is confirmed once the deposit is paid through this link.Booking held, deposit link sent, owner notified08:12Paid. See you in July!Thank you, Anna. Your booking is confirmed.Checkout · 15 JulyThank you for staying with us. If you have a minute, a Google review helps us a lot.
Vila Lazur and Anna are made up. The steps are the ones our hotel systems run.

Every answer in that conversation comes from what the owner gave the system, such as the rooms, what is included and the house rules. Availability is read through the booking connection, and the booking is written into the calendar or booking tool the property already uses.

Hotel AI, for the booking-site inbox

Many guest questions never touch WhatsApp. They arrive through Booking.com, Airbnb and Expedia, each with its own format and no shared place to reply. For that we built Hotel AI, and it is stricter than the WhatsApp conversation above.

Hotel AI connects to the property’s own reservations mailbox, picks out each new guest message, searches that hotel’s knowledge for the answer and replies in the right thread. It sends through the hotel’s own address, so each platform threads the reply correctly. Each hotel’s knowledge, tone and limits are kept separate.

  • It never quotes a price or a vacancy. Guardrails block any invented price or availability. It collects the dates and hands off to staff.
  • It can run in draft first. A dry-run mode writes every reply without sending, so a hotel can pilot it before trusting it.
  • It works the mailbox you already have. The platforms offer no usable API at this size, so the mailbox is the integration.

The build is written up in Hotel AI, and the story behind it in four inboxes and nowhere to reply.

iButler, for requests during the stay

iButler is a digital concierge. A guest scans the QR code in the room and lands in the hotel’s branded ordering portal, with no app to install and no account to create. They order in their own language from a menu the hotel wrote once. The request reaches the owner’s dashboard and phone as it happens, logged and timestamped, and staff work it through to done. A hotel can go from nothing to a live, multi-language portal in under ten minutes.

Read more in iButler and the hotel app nobody was ever going to download.

When a hotel chatbot is not worth it

  • You get a few enquiries a week. The free WhatsApp Business app with quick replies and a complete Google Business Profile will cover it.
  • Every stay is priced by hand. If each enquiry needs your own judgement on price, the system can gather the dates and the number of guests, and you still close it.
  • Messages already get answered within the hour. Then there is little to recover, even in season.
Two numbers decide itHow many guest messages waited overnight last week, and what one booking is worth to you. Multiply them, and you know what the gap costs each week.

What it costs and how it starts

Our AI systems start from 2,000 euros for one system on your own information, one channel and the booking connection. The price rises with the number of tools it talks to, how strict the rules are, and whether it answers voice calls. The first channel is usually live in about two weeks, and nothing in your current setup has to change. You leave the first meeting with a scope, a price and a start date.

The hotel version is described on Rhomn Systems for hotels and apartments.

Questions owners ask

Does it answer Booking.com and Airbnb messages?

Hotel AI does, through the property’s own reservations mailbox, replying from the hotel’s address so each platform threads it correctly.

Which languages does it answer in?

The guest’s own. It replies in the language the guest wrote in.

Can it promise a room that is not free?

It answers only from what you gave it. In Hotel AI, guardrails block any price or availability it has not been given, and it hands the dates to staff.

Do I have to change my booking tool or channel manager?

No. It works with the tools you already use.

Can I read what it tells guests?

Yes. Every conversation is there for you to read, and it can be switched off in a day.

Find out if it pays for your propertyThirty minutes on a call. Tell us where guest messages wait today, and we will tell you whether a system would pay for itself.Book 30 minutesHotels and apartments

Share this post:

LinkedInX

Journal

Explore More Posts