AI agents for hospitality

Your teams don't search anymore. They ask.

Q: Is our data used to train AI models?

No. Conversations are retained only to improve Max and Ton, and nothing is used to train external models.

Holiday park and campsite teams ask Max in their own words and get answers from live reservations, rates and configuration. Ton does the same for your guests.

Book a demo

Built-in Inherits each user's rights Data stays in Europe Never used to train external models

Max is in use at

Data

It knows your operations

Reservations, guests, rates and figures, read straight from the platform. Not a chat window on a help centre that only knows what the manual says.

Role

It knows who is asking

Max inherits the rights of the person asking. A receptionist and a revenue manager can ask the same question and correctly get different answers.

Context

It knows where you are

Max knows which module and screen you are working in, so "how do I do this here?" gets an answer for that exact screen.

Eight questions a day, all to the same colleague

A field worker's day at a park that runs on software people have to learn first.

08:05 “How do I move a check-in date again?”
08:40 “The guest at the desk is early and his unit isn't ready. What else is free and clean?”
09:55 “Can you look up the Jansen reservation? He says he paid the deposit.”
11:20 “Who changed this reservation yesterday?”
13:10 “Can a dog stay in this accommodation type?”
14:30 “How many arrivals do we have on Saturday?”
15:45 “How do I raise a work order for a broken light?”
17:20 “Why did the price on this booking change?” Back office replies tomorrow morning.

Every one of these has an answer. It sits in the platform, or in the head of the colleague who has run the back office for years. The problem is not that your people ask too many questions. Your software only answers questions when people already know where to find the answer.

Max takes the question as they would put it to a colleague, finds the answer in your own data and configuration, and shows them where to act. At De Krim Texel, seven of those eight questions no longer reach the back office.

Max · for your teamsIn pilot

Ask it. Find it. Calculate it. Write it. Act on it

Five things your staff do all day, now done by asking.

Get answers

Step by step, for the module you are working in, with the source cited so you can check it or tell us the documentation is wrong.

How do I move a check-in date?

Find information

Max finds the reservation and gives you what you need: travel party, channel, who changed what and when. Lists as well as single records, without navigating four screens to reach one field.

Which bookings have the late check-out add-on?

Get insights

Max calculates the answer from your live figures, not from last month's export.

Occupancy per accommodation type in August versus last year.

Create content

A conversation becomes a clean support ticket, a work instruction is shortened for a colleague. Pilot users name this one first, because it ends the back-and-forth.

Write this up for support.

Take action

Max prepares changes to reservations, work orders or rates, and explains the consequences before anything moves. With your confirmation, Max executes the action.

Prepare a rate change for these dates.

Ready. 27 units affected from 12 to 19 July. Confirm?

AI that is already making a difference

For me, the value of AI lies in making daily work easier, so our staff have more time and attention for our guests.

Jolanda Kamphuis Commercial Director, De Krim Texel

7 of 8

questions a day resolved without the back office

Questions a day from field workers at De Krim Texel now resolved without the back office.

6 weeks → 1 hour

to work independently

For a new employee at De Krim Texel to work independently.

Figures from De Krim Texel, where Max went live first. Libéma followed as the second pilot operator.

One AI assistant for every role at the park

Built from real questions asked at real front desks, contact centres and back offices, and tested with them before release.

Reception

“He's early and his unit isn't ready.”

Max finds one that is clean, same type, same dates, and shows where to switch it.

Contact centre

“A group of 15 next weekend, one accessible unit.”

Max returns what is free on those dates.

Park manager

“Occupancy per accommodation type, August against last year.”

On live figures, in one sentence.

Revenue

“What happens if I change this price rule?”

Max shows which dates, accommodation types and channels it touches before you act.

Configuration

“How do I make this accommodation type bookable?”

The steps, in the module you are in, with the source cited.

Back office

“Write this up for support.”

A complete ticket, so the first reply is a fix rather than a question.

In the guest help desk · In pilot

Ton drafts. Your team sends.

Ton reads the guest's message and the reservation behind it, writes a suggested reply and hands it to whoever is on the inbox. Nothing reaches a guest unreviewed.

On your website · In preview

A chat widget, installed in about a minute

Ton answers guests in real time on your own site. A colleague can take over mid-conversation while Ton keeps reading along, and the conversation closes with a summary.

Most AI assistants read the manual. Our AI agents read your parks

What a third-party agent cannot do with your documentation alone.

It sits where the data lives

The reservation, the rate rule, the allotment and the configuration, plus the Help Centre. A third-party agent can read your manual. It cannot tell you why this reservation's price moved.

It knows how you work

A context layer, built with each operator, means answers reflect your own use of Maxxton rather than generic best practice.

It asks back

When a question is incomplete, Max asks which reservation you mean rather than guessing, and holds the thread through a multi-step configuration.

It speaks your team's language

Staff ask in their own language and get the answer in it, from one shared knowledge base. No internal documentation to translate per country.

It was tested with real roles

Used daily by pilot operators across pricing, configuration, communication, and the front desk, up to management. Its scope comes from their questions, not our assumptions.

A team improves it every week

We score every conversation and classify every gap. A dedicated AI team in Amsterdam builds what the agents learn they cannot yet do.

Every question your team asks improves the platform

The agents also show us where your people struggle with the software, question by question.

Knowledge gap

The answer exists, the agent had not learned it. A new Help Centre article, drafted from the conversation and approved by a person, closes it within days.

Skill gap

The platform can do it, the agent cannot reach it yet. We connect the agent to that part of the platform.

Feature gap

The platform cannot do it yet. It reaches the product teams as prioritised input from real usage.

An AI agent your IT and data protection officer can sign off on

The roles and rights you already set

Max has no rights of its own. It works within the roles set in user access management: view, create, edit, delete or execute per application, narrowed by location, distribution channel or cash account, and the GDPR access level on the role. No separate AI permissions to maintain.

A person confirms every change

Max reads and explains freely. Anything that changes data is prepared by Max and executed only with your confirmation.

Your data stays in Europe

Processed and stored in Europe on Google Cloud, inside the same environment and data protection as the rest of the platform.

Not used to train external models

Conversations are retained only to improve Max and Ton. Nothing is used to train external models.

What operators ask before they switch it on

Are Max and Ton chatbots?

No. A chatbot reads documentation. Max reads your live reservations, rates, availability and configuration inside the platform, within the rights of the person asking. Ton answers guests from your own park information and the reservation behind their question.

What is the difference between an AI agent and an AI assistant?

An AI assistant answers from what it has been given to read. An AI agent works inside the platform: it looks up live data, calculates and, with your confirmation, acts. Max and Ton are conversational AI agents: you ask in your own words, and they work on your live data.

Can Max change data in our system?

Yes, but Max first prepares the reservation, work order or rate change and explains what it will affect for you to review and confirm. Nothing is executed until the person asking confirms, and Max can never do anything that person could not do themselves.

Who can see our data, and where is it processed?

Only users who could already see it. Max works within the roles, rights and GDPR access levels you already set in user access management, and has no broader view of its own. Data is processed and stored in Europe on Google Cloud.

Is our data used to train AI models?

No. We retain conversations only to improve Max and Ton, and we don't use them to train external models.

Does Ton talk to our guests directly?

No, not in Guest Help Desk: Ton drafts the reply and your team reviews and sends it. On your website, Ton answers guests in real time through a chat widget that is in preview and not yet live for operators. A colleague can take over any conversation.

Do new staff still need training on the software?

Far less. At De Krim Texel, it took a new employee about six weeks to work independently; now it takes roughly an hour because they ask Max instead of learning every screen first.