AI Travel Planner Integration

My React Native travel-booking app (Android & iOS) already handles user accounts, payments and our proprietary flight + hotel APIs. Now I want to layer an itinerary engine on top of it that combines generative AI with human editing. What I expect the module to do • Accept the traveller’s dates, budget, interests and travel style, then generate a full, day-by-day plan across one or several destinations. • Let the user tweak anything—swap hotels, shift dates, add or drop cities, change activities—without leaving the flow. • Offer direct in-app booking for flights, hotels, transfers, local transport and activities; all powered by the APIs already live in the codebase. • Whenever a date, destination or activity changes, instantly refresh prices, availability and the overall trip total, then surface the delta to the user. • Keep a running budget tally and present a clean, shareable trip summary once the traveller commits. Tech & integration notes The existing stack is React Native with Redux and TypeScript; back-end endpoints return JSON for flights, hotels and activities. You are free to choose the AI layer (OpenAI, Claude, etc.) as long as the prompt can be tuned to respect budget, interests and travel style, and the response can be normalised into our itinerary schema. Persistence is via Firebase; new collections will be required for draft trips and AI metadata. Acceptance criteria 1. A user can request an itinerary and receive a structured plan in <10 s. 2. Manual edits accurately propagate through the schedule and cost calculator. 3. All bookings triggered from the itinerary hit existing endpoints with no breaking changes. 4. Unit tests cover critical reducers, price-refresh logic and API calls. 5. The feature passes QA on both Android (Pixel 6) and iOS (iPhone 14) builds in TestFlight. Please share comparable projects you have delivered, the toolset you propose for the AI layer, a timeline with milestones, and a rough cost estimate.