Hospitality & Tourism · Worked scenario
“Guests message us on six channels, and the GM sees none of it until the review is posted.”
AI agents and automation for hotels and hospitality in Dubai, the UAE and the GCC. Requests arrive through WhatsApp, OTA inboxes, email, phone and the desk, against Opera and SynXis. Group enquiries sit in personal inboxes over the weekend. Feedback surfaces when it is public, where it sets next season’s rate.
What it costs you today
- 11.2% more a hotel can charge at the same occupancy for a one-point rise in its review score on Travelocity’s five-point scale, in a 2012 Cornell University study¹
- 23.6 hours was the median wait for a person’s reply to a private-event enquiry, among those that got one, and 19% never got one, in a US vendor study of 807 enquiries to restaurants, bars and event venues²
- More than 35% is how far Dubai’s revenue per available room fell in the first half of 2026, with occupancy at 56.4%, in JLL figures reported by The National⁴
In a market this full, the margin is in the response.
A one-point rise in review score on Travelocity’s five-point scale lets a hotel charge 11.2% more at the same occupancy, in a 2012 Cornell University study¹. In a US vendor study of 807 private-event enquiries to restaurants, bars and event venues, 19% never got a reply from a person, and those that did waited a median 23.6 hours². Dubai had 154,264 hotel rooms at 80.7% average occupancy in 2025, a third record year with 19.59 million international visitors³; in the first half of 2026 revenue per available room fell more than 35%, with occupancy at 56.4%, in JLL figures reported by The National⁴. In a market this tight, the margin is in the response.
○ guest request · ◇ enquiry
What they tried
A chatbot on the website. Guests kept messaging WhatsApp, and the chatbot could not see the PMS.
What we build
The First System: a guest-messaging concierge on WhatsApp connected to the PMS, answering in the guest’s language and routing requests to the department that owns them. It grows into group-enquiry capture with drafted proposals, in-stay sentiment with recovery before the review, direct ordering off the aggregators, and the duty manager’s live view.
What a Tuesday looks like
- Three days before arrival
A transfer is booked and a dietary note reaches F&B.
- Saturday 11:20
A wedding enquiry becomes a drafted proposal on the pipeline board within minutes.
- Day two of a stay
A guest’s tone sours over noise. Recovery is triggered while they are on property.
- 19:00
Diners book by message and outlet managers see covers by hour.
What changes
The guest thread, start to finish
Before Requests on six channels, invisible to the GM.
After An AI concierge answers in the guest’s language and routes to the right department.
Group and event enquiries
Before A wedding enquiry sits in a salesperson’s inbox.
After Every enquiry captured, the proposal drafted, stalled ones flagged.
Sentiment before the review
Before Problems surface on review sites.
After In-stay tone read and recovery triggered while the guest is still on property.
Restaurants by conversation
Before Reservations by message, handled by hand.
After Reservations, deposits and changes taken by an agent.
Every agent we deploy says it is an AI, and a person can always take over.
What we measure
Time to first reply by channel, group-enquiry response time, review score, direct-booking share, requests resolved per department.
What this needs from you
- PMS, OTA inbox and WhatsApp Business access.
- Rate and space data.
- A department owner per request type.
Delivered through
Questions a careful buyer asks
- What do you build first?
- The First System: a guest-messaging concierge on WhatsApp connected to the PMS, answering in the guest’s language and routing requests to the department that owns them.
- What does it need from us?
- PMS, OTA inbox and WhatsApp Business access. Rate and space data. A department owner per request type.
- What do you measure?
- Time to first reply by channel, group-enquiry response time, review score, direct-booking share, requests resolved per department.
Every guest heard on the channel they chose, and every department seeing it at once.
Tell us about your business Talk to an engineer
Our AI consultant gathers the details and passes a report to one of our engineers, who replies by email, WhatsApp or phone.
Other sectors
Sources
- ¹ Cornell Center for Hospitality Research (Anderson), via HospitalityNet: the impact of social media on lodging performance, 2012
- ² Hermetic AI: 2026 hospitality lead-response benchmarks, 807 mystery event enquiries to US venues (a software vendor’s study)
- ³ Dubai Media Office: Dubai’s tourism industry achieves a third successive record-breaking year, February 2026
- ⁴ The National, citing JLL: Abu Dhabi tops UAE occupancy rates in the first half of 2026 as Dubai’s revenue per available room falls, August 2026
Figures from the sources cited, checked on 6 November 2026. Studies from outside the UAE are named as such.
A composite built from the published figures cited and the systems this sector runs on, not a named client.