TripMate

TripMate — AI Multi-Agent Travel Planner Planning a trip usually means juggling five different tools — destination research, budgeting, hotel comparison, route mapping, and daily planning. TripMate replaces all of that with one coordinated AI experience. The user describes their trip: destination type (mountains, beach, city, adventure, historical), starting location, travelers, trip length, and budget. TripMate's Destination Agent — powered by Gemini AI — suggests four real, specific destinations, each with a match score, estimated cost range, and why it fits. Once a destination is chosen, TripMate assembles a complete plan through specialized responsibilities that work like a virtual AI travel team: Budget Agent calculates transport, accommodation, food, activities, and buffer costs using editable local logic, updating instantly as inputs change. Hotel Agent surfaces nearby stays with pricing, ratings, and amenities. Route Agent builds a visual route with waypoints and stops. Itinerary Agent — the second live AI call — generates a realistic day-by-day schedule that accounts for travel time and rest, avoiding overcrowded plans. Checklist Agent produces a destination-specific packing list. Only two actions trigger a live AI call — destination suggestion and itinerary generation — both cached for the session. Everything else runs on fast, deterministic local logic, keeping the app reliable and cost-efficient. AI failures fall back gracefully instead of breaking the experience. Visually, TripMate uses an immersive travel aesthetic — a mountain-dusk hero, glassmorphic cards, and warm amber accents against alpine greens — aiming to feel like a real, launch-ready product. Built end-to-end on native.builder, with Supabase managing the Gemini API key and Edge Function execution. No login required — trip data lives in local session state.