Experiences Vertical

Bookable experiences and activities (tours, private guides, excursions, classes).

LOE: ~68-98 hours

---

Epic: Experiences Data Acquisition

Plan: Build crawlers for Viator, GetYourGuide, and Klook to acquire bookable experience data including availability, pricing, and ratings.

Architectural Context: Each source covers different geographies and experience types. Viator is global, GetYourGuide is strong in Europe, Klook focuses on Asia-Pacific. Crawlers run in parallel with partner API authentication handling. Results are deduplicated across sources.

Tasks

• Build Viator crawler
• Build GetYourGuide crawler
• Build Klook crawler
• Handle partner API authentication requirements

---

Epic: Experiences Schema and Storage

Plan: Design the experience data model, implement SQLite storage with FTS5 search, and build MCP tools for agent consumption.

Architectural Context: The data model captures type (tour, excursion, class), duration, price, rating, location, and availability. SQLite provides embedded storage with FTS5 for full-text search. MCP tools expose experience data to Research Agents.

Tasks

• Design experience data model (type, duration, price, rating, location)
• Implement SQLite tables and FTS5 search
• Build experiences-specific MCP tools

---

Epic: Experiences Booking Integration

Plan: Implement real-time availability checking and build a booking flow that hands off to Atlantes.

Architectural Context: Availability checking queries partner APIs for real-time booking windows. The booking flow collects advisor and member context, then hands off to Atlantes for reservation creation. This integrates the Intelligence layer with the Integration layer's hand-off flow.

Tasks

• Implement availability checking
• Build booking flow integration with hand-off to Atlantes