Agentrolux

Cet exemple est en cours de mise à jour. Le texte ci-dessous est encore en anglais.

Example · HospitalityGuest enquiry and arrival agent

The question at 23:10, answered from the record

An agent that answers the pre-stay questions guests actually ask, takes booking enquiries for the rooms you have open, and sends arrival details from the property record.

Illustrative scenario — not a real client result

Guest messages · 23:10

Simulated example
  • 7 guest messages answered

    23:11 · Check-in time, parking, cot — from the property record

    Answered
  • Arrival instructions sent

    23:14 · Door code released 24 hours before arrival

    Answered
  • “Can we get a late checkout?”

    23:16 · Money and housekeeping · with a person

    Person

No amenity or price is ever guessed

Guest enquiry and arrival agent · Hospitality

  • Sourced facts
  • A person decides
  • Property record only
01 — A typical day

Three small failures that add up.

What hospitality owners tell us goes wrong in an ordinary week.

The same six questions, all evening

Check-in time, parking, cots, dogs, late checkout, the door code. Each one interrupts the person on shift.

See the sector picture

An amenity that does not exist

“Yes, there's parking” — because a general model wanted to be helpful. The guest arrives with a car and nowhere to put it.

See the record rule

When the booking system is down

A supplier outage, and suddenly there is no way for a guest to ask anything or confirm a stay.

See the outage cases
01 — A typical day · the evidence

What the numbers actually say.

Each figure opens its source, year, sample and limits. Click a badge to check it.

01
Source

Hotel AI adoption survey 2025 (Schegg, HES-SO), via MyRent summary

2025

Method and sample are recorded in the source registry.

Read the source for its scope and limits.

Open the source

41%

of surveyed hotels reported using AI

1,485 hotels in six European countries, surveyed January–April 2025 (Schegg). Reported through a vendor summary, not the original paper.

Hotel AI adoption survey 2025 (Schegg · 2025
02
Source

Eurostat, use of artificial intelligence in enterprises

2025

Method and sample are recorded in the source registry.

Read the source for its scope and limits.

Open the source

58.82%

of EU accommodation firms using AI use it for marketing or sales

Eurostat ICT survey, early 2025: 157,000 enterprises sampled. Covers firms with 10+ staff, so it excludes the smallest guesthouses.

Eurostat · 2025
03
Source

Revinate hospitality benchmark, summarised by the Irish Hotels Federation

2024

Method and sample are recorded in the source registry.

Read the source for its scope and limits.

Open the source

5.1

inbound lead calls per room in a hospitality benchmark

Revinate benchmark, 4.6 million calls analysed. Vendor data, not an independent survey.

Revinate hospitality benchmark · 2024
02 — What the helper does

From your records to a decision you own.

The agent answers pre-stay questions from the property record: check-in and check-out times, access, parking, cots, pets, breakfast and the route from the station.

It takes booking enquiries for the rooms and dates the property has opened, sends the confirmation, and releases the door code at the moment you choose.

Money, complaints, refunds and anything about a guest's safety go to a person. If a field is missing from the record, the agent says so instead of filling it in.

6 steps · 2 decisions you own

Text version of the flow
  1. 1. Read the message — Which stay, which question.
  2. 2. Answer from the record — Times, access, amenities.
  3. 3. Take the booking enquiry — Only open dates.
  4. 4. Send arrival detail — Code, route, rules. Human checkpoint: You set the access rules.
  5. 5. Stop at money and complaints — Refunds, changes, damage. Human checkpoint: A person decides anything with money.
  6. 6. Close the loop — One thread, one record.
02 — What the helper does · a simulated example

One simulated example, step by step.

Use the buttons or the left and right arrow keys. Nothing here is a real customer.

Simulated example

Saltings House, four rooms

Guest messaging and email · 23:10

1 / 6

A four-room guesthouse on the coast. The agent answers guests outside reception hours. Every message is simulated; no real guest data is used.

  1. This is Saltings House's assistant — I can answer questions about your stay and send arrival details. Reply HUMAN for a person.

03 — What you decide

The helper prepares. You decide.

The agent answers the guest. The owner decides the stay.

  • The record is the only source

    Every answer names the field it came from; an empty field is reported as unconfirmed.

  • Access on your terms

    The code goes out when you say, after the conditions you set are met.

  • A person takes over

    Refunds, discounts, complaints, damage and safety concerns go to the named contact.

  • You can pause it

    Pause per property or entirely; messages fall back to your own inbox with an away reply you wrote.

Without your approval · No amenity is invented, no price is changed, no refund is agreed, and the door code is released only on your schedule.

04 — What changes when a safeguard is missing

Documented cases, not scare stories.

Three documented cases with their sources. Two concern systems failing; one concerns a business disappearing mid-stay. Each shapes a fallback on this page.

2025

Sonder

Guests told to leave with a day's notice

After Marriott ended its licensing agreement, Sonder announced an immediate wind-down and planned liquidation. Guests reported being told to leave in less than 24 hours; the interim CEO attributed problems partly to integration costs and revenue decline.

What the helper would do differently

The page promises a published continuity plan: if the property cannot honour a stay, the agent stops taking enquiries and a person contacts every guest with a relocation route — not an automated message.

< 24 hrs

notice given to guests before wind-down

2024

Omni Hotels

A cyberattack took the door locks with it

After a cyberattack, Omni shut down systems, initially affecting reservations, room-door locks and point-of-sale across the chain.

What the helper would do differently

The agent is never the only channel or the only key. A tested offline check-in, a physical key route and a phone number are part of the setup, not an afterthought.

Chain-wide

reservations, locks and tills affected

2026

NOHGA HOTEL

Booking system down, no estimate

A server failure in the hotel's external reservation system made online accommodation bookings temporarily unavailable. The hotel initially gave no restoration estimate, then reported the issue resolved.

What the helper would do differently

Booking enquiries have an alternative route — a phone number and a direct email — printed in every message the agent sends, so a supplier outage is not a closed door.

Online bookings

unavailable until the supplier fixed the server

04 — What changes · the evidence

The rest of the evidence.

Every figure resolves to a source with its sample and limits. No figure without one.

  • 41%

    of surveyed hotels reported using AI

    1,485 hotels, six countries, January–April 2025. Reported via a vendor summary; the sample is not representative of every European hotel.

    Source

    Hotel AI adoption survey 2025 (Schegg, HES-SO), via MyRent summary

    2025

    Method and sample are recorded in the source registry.

    Read the source for its scope and limits.

    Open the source
  • 1,485

    hotels in the European survey

    Online survey by Schegg, HES-SO. Italian responses came mainly from one region.

    Source

    Hotel AI adoption survey 2025 (Schegg, HES-SO), via MyRent summary

    2025

    Method and sample are recorded in the source registry.

    Read the source for its scope and limits.

    Open the source
  • 58.82%

    of EU accommodation AI users apply it to marketing or sales

    Eurostat, early 2025. A purpose-of-use share among AI users, not overall adoption.

    Source

    Eurostat, use of artificial intelligence in enterprises

    2025

    Method and sample are recorded in the source registry.

    Read the source for its scope and limits.

    Open the source
04 — What changes · your week

The same week, two ways.

Drag the handle, or use the left and right arrow keys when it is focused.

Typical week, beforeTypical week, after
  • Mon

    Eleven messages overnight; the owner answers them at breakfast.

  • Tue

    A parking question is answered from memory and is wrong.

  • Wed

    The door code is texted to the wrong guest, then to the right one.

  • Thu

    A refund request sits in the inbox while the guest waits at the door.

  • Mon 23:11

    Overnight messages answered from the record, each with its source field.

  • Tue

    Unlisted answers are flagged to the owner instead of guessed.

  • Wed

    Codes release on schedule, to the booking they belong to.

  • Thu

    Refunds and changes are with the owner, and the guest knows who will reply.

Illustrative example · not a measured result

05 — Your version

Four questions about your own week.

Keep your answers. They are the brief for the first version we would build for you.

  1. How many guest messages arrive before a stay, and what do they usually ask?

  2. Which details about the property do you want sent automatically, and which not?

  3. What is the booking rule for the rooms you have open this month?

  4. Which request always needs a person: a complaint, an allergy, or a special date?

Next

Put your own figures into the calculator below. It runs in your browser, and nothing is sent or stored.

Put your week into the calculator

Bring the four answers to the free one-hour call. We will use them to sketch the first version, and you keep the written plan either way.

Talk about your business
05 — Your version · your numbers

Put your own week into it.

An estimate, not a promise. The model and its assumptions are printed below.

Your inputs

All channels together.

count

Before, during and after the stay.

count

Reading, checking, replying.

min

Whoever covers reception.

per hour

When nobody is at the desk.

%

Example defaults are typical for a small business. Replace them with yours. Nothing is stored, sent or shared.

Estimated result

Illustrative

Front-desk hours back each month

0–0

Time released at your rate
£0–£0
Messages answered overnight
0–0%
Time to a reply
Already immediate
How we calculated this

Routine work handled by the helper: 70% of the share you marked routine.

Time saved per item: 80% of the minutes you entered.

Monthly volume: weekly figures × 4.33; monthly figures used as entered.

Cost: hours saved × your staff cost per hour, in pounds as an illustration.

Reply time: routine items move within the hour while the helper runs. Complex work still waits for a person.

Questions

The things people ask next.

Only from the property record. If parking, a cot or a pet policy is not recorded, it says so and asks you — it does not fill the gap.

Start here

Show us a week of guest messages.

One hour with last week's messages and your property record. We will show you what the agent would answer, and what it would refuse to guess.