Inspiration# 🧠 Inspiration

Exploring new places should feel personal — not generic. We often rely on popular lists or random suggestions that don’t truly reflect who we are.

TasteTrip AI was created to recommend locations that align with your unique cultural preferences — from the food you enjoy to the music you vibe with — and seamlessly guide you there using intelligent, contextual map-based visuals.

💡 What it does

TasteTrip AI recommends restaurants and destinations tailored to your individual cultural taste — and enhances the discovery experience with:

A taste-based prompt system (e.g., “fantasy novels, lofi music, sushi”)
Qloo’s Taste AI translates your preferences into cultural categories
A smart agent selects and explains fitting nearby venues
Google Maps provides:
- Directions
- Local weather overlay
- 3D view mode
- Rich insights (e.g., terrain, satellite view, neighborhood dynamics)
The system remembers your likes using RAG (Retrieval-Augmented Generation)

🛠 How we built it

Frontend: React + TailwindCSS
Backend: Node.js + Express
Qloo Taste AI: Maps user inputs to multidomain cultural data
Google Maps JavaScript API:
- Interactive maps with 3D view and data layers
- Directions rendering and routing logic
- Dynamic weather and terrain visualizations
OpenAI: GPT for prompt interpretation, explanations, and RAG memory
Supabase: Handles user embeddings, session persistence, and vector search
RAG architecture: Supports personalized memory and evolution over time

🚧 Challenges we ran into

Mapping cultural tastes to geospatially relevant data
Integrating multiple APIs smoothly
Making AI explain why a place matches a user’s unique profile
Managing real-time UX for geolocation and dynamic overlays

🏆 Accomplishments that we're proud of

Created a pipeline that links abstract taste to real-world locations
Integrated AI explanations into map-based suggestions
Delivered weather-aware, 3D-enabled discovery
Engineered persistent user memory and preference learning

📚 What we learned

Taste can be a strong filter for real-world discovery
LLMs paired with cultural embeddings enable deeper personalization
Enhancing maps with live weather, terrain, and interactivity increases decision confidence
RAG architecture brings continuity to user experience