🌍 AI Travel Planner

An intelligent, adaptive travel planning system powered by multi-agent AI orchestration, real-time data integration, and personalized recommendations.

🧠 Overview

This system transforms fragmented travel planning into an intelligent, dynamic experience using:

• Llama-based Orchestrator for reasoning & planning
• GNN Agent for personalized recommendations
• Real-time API Integration for flights, hotels, and weather
• Budget Optimizer for cost vs. comfort tradeoffs
• Streamlit UI for dynamic user interaction
• PDF Export for downloadable itineraries

🏗️ Architecture

ai-travel-planner/
├── backend/               # Core AI agents and orchestration
│   ├── orchestrator.py   # Llama model for intent parsing
│   ├── api_manager.py    # Real-time/mock API integration
│   ├── personalization_gnn.py  # GNN-based personalization
│   ├── budget_optimizer.py     # Cost optimization
│   ├── itinerary_agent.py      # Itinerary generation + PDF
│   ├── utils/            # Utilities and configuration
│   └── main.py           # Unified orchestrator entry
│
└── frontend/             # Streamlit web interface
    ├── app.py           # Main Streamlit app
    ├── components/      # UI components
    └── styles/          # Custom CSS and theming

🚀 Quick Start

Easy Way (Recommended) 🎯

# Just run this - it does everything!
./run.sh

Then choose option 1 for the Web UI.

Manual Setup

Step 1: Install Dependencies

pip install -r requirements.txt

Step 2: Set Up API Keys (REQUIRED for flight and hotel data)

Create a .env file:

# REQUIRED for flights and hotels
SERPAPI_API_KEY=your_key_here  # Get from serpapi.com (100 searches/month free)

# Optional for weather
OPENWEATHER_API_KEY=your_key_here  # Free from openweathermap.org

# Optional for activities
YELP_API_KEY=your_key_here  # Free from yelp.com/developers

Step 3: Run the Application

# Web UI (Recommended)
streamlit run frontend/app.py

# Or test backend directly
python backend/main.py

📚 Detailed Setup Guides

• INSTALL.txt - Quick visual install guide (start here!)
• SETUP.md - Comprehensive setup documentation
• API_SETUP_GUIDE.md - API configuration details
• QUICKSTART.md - Quick reference guide

🧪 Verify Installation

# Test SERP API integration (flights + hotels)
python test_serpapi.py

# Test other APIs
python tests/test_apis.py

# Test workflow
python tests/test_workflow.py

# Test new features  
python tests/test_new_features.py

✨ Features

User Input

• Destination selection
• Date range picker
• Budget slider
• Preference tags (Adventure, Luxury, Nature, etc.)
• Travel group size

Live Processing

• Real-time agent progress updates
• Multi-agent orchestration visualization
• Animated loading states

Results Dashboard

• ✈️ Flight recommendations with airline logos, real-time pricing, and carbon emissions
• 🏨 Hotel options with images, ratings, reviews, and amenities
• 🌤️ Weather forecasts for each day
• 🗺️ Interactive map integration
• 💰 Budget breakdown with detailed cost allocation
• 📥 PDF itinerary download
• 🗓️ Calendar export (.ics)

Feedback System

• User satisfaction ratings
• Itinerary accuracy feedback
• Continuous improvement analytics

🎨 Design Philosophy

• Theme: Deep blue + coral accents
• Layout: Responsive grid-based design
• Animations: Smooth transitions and loading states
• UX: Intuitive, modern, and accessible

🧩 Backend Modules

Orchestrator (orchestrator.py)

• Parses user intent using Llama model
• Validates constraints
• Coordinates agent workflow

API Manager (api_manager.py)

• SERP API Google Flights - Real-time flight data with airline logos
• SERP API Google Hotels - Hotel data with images and ratings
• OpenWeatherMap - Weather forecasts
• Budget-based filtering - Filters results by budget constraints

Personalization GNN (personalization_gnn.py)

• Graph Neural Network for user preferences
• Historical behavior analysis
• Contextual recommendations

Budget Optimizer (budget_optimizer.py)

• Pareto optimization for cost vs. comfort
• Multiple itinerary options
• Value scoring

Itinerary Agent (itinerary_agent.py)

• Consolidates all recommendations
• Generates PDF itineraries
• Exports calendar events

📋 Requirements

• Python 3.9+
• PyTorch
• PyTorch Geometric
• Streamlit
• Transformers (Hugging Face)
• ReportLab
• Other dependencies in requirements.txt

🔧 Configuration

Edit backend/utils/config.yaml to customize:

• API keys and endpoints
• Model configurations
• Budget constraints
• Personalization parameters

📝 Testing

Sample input is provided in backend/utils/sample_input.json for testing the backend pipeline independently.

🤝 Contributing

This is a demonstration project showcasing AI orchestration and real-time travel planning capabilities.

📄 License

MIT License