AI Concierge for Marriott Codefest

Theme: Booking to Belonging

An AI-powered travel concierge web application that personalizes the travel experience by analyzing user preferences through photo analysis and intelligent booking recommendations.

🎯 Project Overview

This application helps travelers discover their travel personality and get personalized recommendations for flights, hotels, and attractions.

Key Features

1. **Photo Analysis

   • Upload travel photos
   • AI-powered analysis using Google Gemini Vision API
   • EXIF metadata extraction (location, date, camera info)
   • Travel personality profiling
   • Personalized hotel recommendations

2. **Auto Booking System

   • Flight search and booking via Amadeus API
   • Hotel recommendations and reservations
   • Attraction discovery

🛠️ Tech Stack

Frontend

• Framework: Next.js 14 with TypeScript
• Styling: Tailwind CSS
• UI Theme: Marriott-inspired (red/white/gold palette)
• Features:
  • Drag-and-drop file upload
  • Real-time image preview
  • Responsive design

Backend

• Framework: Flask (Python)
• AI Services:
  • Google Gemini API (image analysis)
  • Mistral AI Voxtral (voice - planned)
  • Amadeus APIs (booking - planned)
• Image Processing: PIL, exifread

📦 Installation & Setup

Prerequisites

• Node.js 18+ and npm
• Python 3.9+
• Google Gemini API key

Backend Setup

1. Navigate to backend folder:

cd backend

2. Create virtual environment:

python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

3. Install dependencies:

pip install -r requirements.txt

4. Configure environment variables:
   • Copy .env.example to .env
   • Add your Google Gemini API key:

GEMINI_API_KEY=your_actual_api_key_here

5. Run the Flask server:

python app.py

Backend will run on http://localhost:5000

Frontend Setup

1. Navigate to frontend folder:

cd frontend

2. Install dependencies:

npm install

3. Configure environment:

   • .env.local file is already created with default backend URL
   • Modify if your backend runs on a different port

4. Run the development server:

npm run dev

Frontend will run on http://localhost:3000

🚀 Usage

1. Image Analysis Flow

1. Open http://localhost:3000 in your browser
2. Click "Discover Your Travel Style" or navigate to /image-analysis
3. Upload 1-10 travel photos (drag & drop or click to select)
4. Click "Analyze My Travel Style"
5. View your personalized travel personality profile with:
   • Travel style (Adventure Seeker, Luxury Traveler, etc.)
   • Personality traits
   • Favorite activities
   • Destination preferences
   • Recommended hotel types

2. Testing the Backend

Test the Gemini API connection:

curl http://localhost:5000/test_gemini

Health check:

curl http://localhost:5000/health

📁 Project Structure

Balloon Muscles/
├── backend/
│   ├── app.py                    # Main Flask application
│   ├── requirements.txt          # Python dependencies
│   ├── .env                      # Environment variables (API keys)
│   ├── .env.example             # Environment template
│   ├── utils/
│   │   ├── gemini_client.py     # Gemini Vision API integration
│   │   └── exif_extractor.py    # EXIF metadata extraction
│   └── uploads/                 # Temporary image storage
│
├── frontend/
│   ├── app/
│   │   ├── page.tsx             # Home page
│   │   ├── layout.tsx           # Root layout
│   │   └── image-analysis/
│   │       └── page.tsx         # Image analysis page
│   ├── components/
│   │   ├── ImageUpload.tsx      # File upload component
│   │   └── AnalysisResults.tsx  # Results display component
│   ├── .env.local               # Frontend environment variables
│   ├── package.json             # Node dependencies
│   └── tailwind.config.ts       # Tailwind configuration
│
└── README.md                     # This file

🎨 Design Philosophy

The UI is inspired by the Marriott website with:

• Color Palette:
  • Primary Red: #8B0000 (Marriott signature)
  • Accent Gold: #B8860B
  • Neutral: Gray scale for backgrounds
• Typography: Clean, modern sans-serif
• Layout: Spacious, luxury-focused, responsive
• Animations: Subtle hover effects and transitions

🔑 API Keys Setup

Google Gemini API

1. Visit Google AI Studio
2. Create a new API key
3. Add to backend/.env: GEMINI_API_KEY=your_key_here

Mistral AI (Future)

1. Visit Mistral AI Platform
2. Generate API key
3. Add to backend/.env: MISTRAL_API_KEY=your_key_here

Amadeus (Future)

1. Visit Amadeus for Developers
2. Create an app to get API key and secret
3. Add to backend/.env:

AMADEUS_API_KEY=your_key_here
AMADEUS_API_SECRET=your_secret_here

🧪 API Endpoints

Backend (Flask)

Endpoint	Method	Description
/health	GET	Health check
/test_gemini	GET	Test Gemini API connection
/analyze_images	POST	Analyze travel photos

POST /analyze_images

Request:

• Content-Type: multipart/form-data
• Body: images (multiple files)

Response:

{
  "success": true,
  "images_analyzed": 3,
  "metadata": [...],
  "analysis": {
    "travel_style": ["Adventure Seeker", "Nature Lover"],
    "activities": ["Hiking", "Photography"],
    "destination_types": ["Mountains", "Forests"],
    "personality_traits": ["Adventurous", "Independent"],
    "recommended_hotels": ["Boutique Hotel", "Mountain Resort"],
    "insights": "This traveler loves outdoor adventures...",
    "confidence_score": 0.85
  }
}

🎯 Next Steps (Roadmap)

• Task 1: Image to Gemini Analysis
• Task 2: Voice + Sentiment Analysis with Mistral Voxtral
• Task 3: Amadeus API Integration for Booking
• Google Drive integration for photo import
• User authentication and profile saving
• Booking history and itinerary management

🐛 Troubleshooting

Backend won't start:

• Ensure Python virtual environment is activated
• Check that all dependencies are installed: pip install -r requirements.txt
• Verify .env file exists with valid API keys

Frontend can't connect to backend:

• Ensure Flask server is running on port 5000
• Check NEXT_PUBLIC_API_URL in frontend/.env.local
• Disable any firewall/CORS blockers

Image analysis fails:

• Verify Gemini API key is valid
• Check image format (PNG, JPG, JPEG, GIF, HEIC)
• Ensure images are under 10MB each
• Check Flask console for error messages

Linter/TypeScript errors:

• Run npm run build to check for type errors
• Install missing types: npm install --save-dev @types/[package-name]

📝 License

Built for educational purposes - Marriott International Codefest at Virginia Tech 2025.

👥 Contributors

Created for the "Booking to Belonging" theme at Marriott Codefest.

---

Happy Traveling! ✈️🏨