Discover Puerto Rico, the island’s official destination marketing organisation, has engaged Satisfi Labs to build an AI platform that matches visitors and meeting planners to local businesses. It is due on DiscoverPuertoRico.com in the first quarter of 2027.
A destination board, not a booking platform, is building the layer that decides which hotels, venues and operators an assistant shows a traveller.
What the platform does
Two agents ship at launch. One serves leisure travellers and matches them to experiences, accommodation and businesses. The other serves meeting and event planners and matches on programme needs such as venues, hotels, logistics and destination services.
Both draw on the island’s verified sources, not the open web. A matching layer routes each request to specific hotels, meeting spaces, attractions and destination management companies.
Matching follows why people travel rather than what they type. PhocusWire reported that the DMO identified 27 descriptors of what people seek from a holiday and grouped them into nine core motivators, with each business assigned one primary and two secondary. Two travellers who type similar queries can be sent to different places.
Businesses must show the right insurance and certifications before the platform recommends them. PhocusWire also reported that about 30% of the island’s businesses are indexed so far, with two team members adding the rest.
Who owns the data
Discover Puerto Rico keeps its knowledge graph, conversation data and intent signals. The graph is built to be “citation-ready” for external AI assistants and travel-planning tools. ChatGPT-style assistants can therefore cite the board’s verified supply, while the board sees what people asked for.
Satisfi Labs serves nearly 800 clients across sports, entertainment and tourism. Its backers include Google, Major League Baseball, Level Equity, Techstars and Florida Funders.
How success is measured
The first measure is qualified referrals to island businesses, tracked in Satisfi Labs’ Agent Performance Console. Later phases link the agents to the board’s marketing channels, so a traveller who clicks an ad lands in a conversation. Interaction data will also be used to spot gaps in tourism supply for economic development.
Why the island is doing this now
In Discover Puerto Rico’s update of 29 October, lodging revenue passed US$1.64 billion through September 2026, up 6.2% and a record for the period. Hotels took US$1.02 billion and short-term rentals US$622 million.
Arrivals moved the other way. San Juan airport counted 5.08 million passengers year-to-date through September, down 3.1%, and the Spirit Airlines closure alone cost 246,000. Revenue is rising on fewer visitors, which raises the value of each one.
The board also cites a Future Partners study: 32.2% of American travellers now use AI for trip planning, against 24% a year earlier, while only 21.1% use official destination websites. If travellers ask an assistant first, the board’s data has to be what the assistant reads.
What it means for APAC
Hotels. In a destination with a verified graph, a property competes on how well its profile is described, not only on rate and review score. Hotels in APAC should expect boards to ask for structured data, certifications and motivator-style tags before listing them in a board-run assistant.
DMCs. The meeting planner agent matches on venues, hotels, logistics and destination services, and DMCs sit among the businesses it connects. A DMC outside the board’s index will not be recommended through it, and insurance and certification checks decide who gets in.
OTAs. Because the graph is built for outside assistants to cite, those assistants get a direct route to verified supply. An OTA that counts on being where discovery starts now has a destination-owned source competing for that moment.
Airlines. The Spirit closure shows how one carrier exit lands in arrival counts. Intent signals from a board-run assistant could show carriers which destinations travellers are asking about.
The Puerto Rico platform arrives in Q1 2027. APAC tourism boards weighing AI plans have a working model to copy or compete with, with a vendor, a taxonomy and a data-ownership stance to study.
Key takeaway
Tourism boards and DMCs in APAC should test this model first, because they will be asked to supply structured data, tags and certifications. Hotel and OTA teams can wait for the Q1 2027 launch, but should start tidying property descriptions and supplier records now. We would watch the referral numbers Puerto Rico reports from the Agent Performance Console, since they will show whether a board-owned assistant sends bookings or only answers questions.
