SEOUL – October 29, 2026 -- Generative AI search and answer services accessed DealLabs' travel database more than 11,000 times per day in mid-October, up from around 600 daily accesses in late September, a nearly 20-fold increase in under a month.
DealLabs' database holds 280,000 verified records across seven cities
As of October 13, the Seoul-based company's database covered Seoul, Busan, Tokyo, Osaka, Bangkok, Taipei and Singapore. Each record carries its source and verification date. The data feeds more than 23,000 guide pages published in formats AI systems can read directly.
Osaka accounts for more than 114,000 records, including over 42,000 medical institutions and pharmacies and more than 30,000 restaurant permits. Taipei holds more than 63,000 records, among them 7,200 station-to-station fares.
Conflicting event dates expose accuracy gaps in travel AI answers
On the same day, public event listings and organizers' own notices gave different end dates for three of 25 pop-up stores in Seoul, by up to 10 days. DealLabs names business hours, closing days, last train times and fares as the areas where AI travel answers are most often inaccurate. It updates its records continuously from official sources.
A single engine runs all seven cities
All seven cities have AI-ready guide files (llms.txt) and data APIs. Seoul also offers a standard OpenAPI specification. Because one common engine runs every city, adding a new one requires only that city's data and sources.
DealLabs is also piloting "Visit Plan Check," which checks AI-generated travel itineraries against official sources daily.
Founder and CEO Young Cho said what matters is "whether the answer is accurate for that city today."
DealLabs targets 10 cities by end-2026 and 30 by 2028
The 2028 target includes New York and Paris.