South Korea-based travel data company DealLabs has built a database of verified local travel information covering seven Asian cities: Seoul, Busan, Tokyo, Osaka, Bangkok, Taipei and Singapore. As of October 13, the database contained more than 280,000 records, each accompanied by a source and verification date. The company says the information powers more than 23,000 guide pages and is published in formats that AI systems can read directly. The project reflects a growing effort to make local information more structured, traceable and usable by AI-powered search, travel assistants and planning tools.
Structured Local Data Targets a Weak Point in Travel AI
Generative AI tools can produce travel recommendations in seconds, but reliable answers depend on more than a capable language model. Details such as a station’s correct exit, a restaurant’s opening hours, a night market’s operating schedule or the last train of the day can change frequently. Outdated information can turn an otherwise useful itinerary into a frustrating experience.
DealLabs is attempting to address that problem by building a structured data layer for local travel information. The Seoul-based company says its database covered more than 280,000 records across seven Asian cities as of October 13, 2026. Every record includes its source and verification date, allowing users and AI systems to identify where information originated and when it was checked. The company says the data supports more than 23,000 travel guide pages.
The coverage includes Seoul and Busan in South Korea; Tokyo and Osaka in Japan; Bangkok in Thailand; Taipei in Taiwan; and Singapore. Rather than focusing exclusively on popular attractions and destination descriptions, the database also addresses practical information travelers need while navigating a city.
From Transit Details to Medical Facilities
The dataset includes subway station exits, train timetables, first and last train times, transport fares, restaurants, tourist attractions and hospitals. It also accounts for location-specific details, including closing days at Singapore’s hawker centres and business hours at Taipei’s night markets.
The concentration of records in certain cities illustrates the project’s emphasis on detailed local information. According to DealLabs, Osaka accounted for more than 114,000 records, including over 42,000 records relating to medical institutions and pharmacies and more than 30,000 restaurant permits. Taipei had more than 63,000 records, including 7,200 station-to-station fare entries.
These examples point to a different approach from travel content built primarily around destination inspiration. A recommendation engine might identify a restaurant or landmark, but a useful travel assistant also needs dependable operational details to help someone reach the location, understand transport costs and determine whether a venue is open.
For AI systems, those details can be difficult to maintain using general web content alone. Local businesses change hours, transport operators update fares and event organizers revise schedules. A model may generate a plausible answer from previously encountered information without knowing whether the underlying detail remains current.
DealLabs says it continuously updates records using official sources. In one example cited in its announcement, the company compared public event listings with organizers’ notices for 25 Seoul pop-up stores. Three listings had conflicting end dates, with discrepancies of up to 10 days. The example illustrates how apparently minor inconsistencies can affect travel plans. PR Newswire
APIs and AI-Readable Formats Expand Access
DealLabs is distributing its data through city guide pages, data APIs and files designed to help AI systems discover relevant information. The company says all seven cities covered by the announcement have AI-readable llms.txt files and data APIs, while Seoul also offers an OpenAPI specification.
These interfaces could make the information easier for external services to retrieve and integrate into travel search, conversational assistants and itinerary-planning products. Instead of relying only on a model’s internal knowledge or a broad search result, an application could query a structured source for a specific local fact.
The approach also separates the data layer from the AI model itself. DealLabs does not need to build a general-purpose foundation model to pursue this market. Its potential role is to maintain specialized, location-specific information that other AI products can use.
The company says its seven city datasets operate on a common engine, allowing additional locations to be added by incorporating their data and sources. That architecture could support expansion, although the quality and maintenance requirements of each new market will remain important factors.
DealLabs also reports growing demand from generative AI search and answer services. According to the company, daily access to its data rose from approximately 600 requests on a seven-day average in late September to more than 11,000 in mid-October. The company describes this as nearly 20-fold growth in less than a month. These are company-reported access figures, not independently verified measures of adoption, accuracy or end-user engagement. PR Newswire
Itinerary Verification Could Become a Separate AI Service
Beyond providing information for AI-generated answers, DealLabs is piloting a service called Visit Plan Check. The tool checks AI-generated travel itineraries against official sources daily, according to the announcement.
That feature addresses a related challenge: even if an itinerary is accurate when created, its details can become stale before the trip. A restaurant may change its hours, a temporary attraction may close, or a transport schedule may be revised. Checking the itinerary against source information could help identify discrepancies before travelers rely on it.
However, the announcement does not establish how widely the checking service is being used, how often it detects errors or whether it can verify every type of itinerary recommendation. Those measures will be important in evaluating its practical value.
DealLabs plans to expand its coverage to 10 cities by the end of 2026 and 30 cities, including New York and Paris, by 2028. Reaching that target will require the company to maintain source quality and freshness across different languages, transport systems, business directories and local regulations. PR Newswire
The broader opportunity is clear: as travel planning becomes more conversational, dependable local data may become an important component of AI infrastructure. Yet a large database alone does not guarantee accurate answers. The long-term test will be whether the information stays current, can be retrieved reliably and is adopted by AI services in ways that meaningfully improve the traveler’s experience.
Market Landscape
Travel AI is expanding from conversational recommendations toward practical assistance, including itinerary creation, transport guidance and local discovery. That shift increases the importance of structured, current information that can be checked against identifiable sources.
DealLabs is positioning itself as a specialized data provider rather than a general AI model developer. Its combination of sourced records, verification dates, city guides and APIs is designed to help AI systems access granular local facts. The model could be useful to travel platforms, AI search providers and itinerary tools, but commercial adoption, data licensing and independently measured accuracy will determine its competitiveness.
Top Insights
- Verified local data: DealLabs says its database exceeded 280,000 records across seven Asian cities as of October 13, with sources and verification dates attached to each record.
- Practical travel coverage: The dataset spans transit exits, schedules, fares, attractions, restaurants and medical facilities, extending beyond conventional destination descriptions.
- AI integration: City-specific APIs and AI-readable files are designed to make structured local information accessible to conversational search and travel-planning systems.
- Rising access: DealLabs reports that AI services accessed its data more than 11,000 times daily in mid-October, up from roughly 600 daily in late September.
- Expansion plans: The company aims to cover 10 cities by the end of 2026 and 30 cities by 2028, a goal that will test its data-maintenance model.
Power Tomorrow’s Intelligence — Build It with TechEdgeAI
