What a hotel front desk sounds like
If you are looking for software and AI for hotels and holiday rentals, you probably don't need another system. You have plenty: a PMS, a channel manager, four or five booking sites, a website, the till, the front desk WhatsApp and a cleaning spreadsheet someone updates when they remember. What's missing is for all of it to talk to each other.
It doesn't matter much whether you run a boutique hotel in the Born, an aparthotel or a building of holiday apartments by the beach. The complaints sound the same:
- “Every booking site gives me a different version of today, and none of them matches the PMS.”
- “Nobody's on at night, so in the morning the phone is full of messages asking about check-in.”
- “Each receptionist gives a different answer to the same question.”
- “To see how the month went I open four systems and a spreadsheet, then reconcile it by hand.”
- “The owners ask for numbers and it takes me days to pull them together.”
None of this gets fixed by switching PMS. It gets fixed by putting something on top.
One screen for your PMS, channel manager and booking sites
We won't ask you to change systems. Your team keeps using the same PMS and channel manager; we build the layer that reads all of them at once.
What you get is one screen: today's arrivals and departures, which rooms or apartments are clean, which site each booking came from, revenue by channel and what's still pending. Up to date, with permissions for each person, on your phone. And the warning arrives before the problem does: three arrivals before 4 pm and one apartment still not cleaned.
When the owners or the fund ask how things are going, you open the same screen. You'll know how the month went without giving up an afternoon to it. That's what systems integration means to us: no more chasing your own data.
What else we build for hotels and holiday rentals
Each piece is a module on the same layer. You start with one and add the next once the first is working.
Guest messages on WhatsApp, answered at night too
Most guest messages are the same few questions: check-in time, parking, the wifi password, whether they can leave their bags. A WhatsApp assistant answers them using the booking details and your house rules, and only hands you what genuinely needs a person. It runs on the same engine as Grippia Recepción, our WhatsApp assistant, set up for your property. There's more on how it works in AI agents for business.
What your team knows, available to everyone
House rules, procedures, suppliers, what to do when the hot water fails. An internal chat that answers from your own documents and shows where each answer came from. So the night shift says the same thing as the morning shift.
A direct-booking website that's part of the system
A booking website only makes sense if it's connected to everything else: direct bookings show up on the same screen as the ones from booking sites, and nobody types them in again. That's why we don't build brochure sites, we build websites with a system behind them.
A real client: SAPO Management, villas and apartments in Lombok
SAPO Management runs villas and apartments in Lombok, Indonesia. Bookings come in through Booking.com, Airbnb, Agoda, Expedia and Traveloka via their channel manager, eZee. In their words:
Our bookings come in through five sites, and each property's information was scattered across all of them.
We built the SAPO Operating System: a layer on top of eZee with one panel for the whole portfolio and a direct-booking website. It replaces nothing: it connects it. It has been underway since October 2026 and isn't finished yet, so there are no figures to share. In November they signed a further piece, an inventory system. The full story is on the SAPO project page, and if your business is in Indonesia, see our work in Lombok and Bali.
Lombok is a long way from Barcelona, but the problem is the same as an aparthotel's here: lots of channels, lots of units and someone to report to.
How we work with hotels: fixed price, in weeks
We start by talking: tell us what's annoying you, not what AI you want. Then we look at the people, the tools and how work flows, from check-in to cleaning. From that you get a proposal: what we'll build, when, and for how much. Fixed, in writing, before we start.
We don't start from scratch: the engines have been running for years, which is why this takes weeks, not months. We ship the smallest useful version, usually the single screen, your team uses it for real, and once it works you decide what comes next.