AI Travel Planner — Smart Budget-Aware Itinerary Generator

An AI-powered travel assistant that creates personalized, budget-optimized day-by-day itineraries with hotels, restaurants, and attractions instantly.

Comment

AI Travel Planner — Smart Budget-Aware Itinerary Generator

Inspiration

Travel planning is often overwhelming: finding hotels, meals, and attractions and keeping everything within budget takes hours of research. I wanted to create a solution that combines AI with real-world travel data to generate fully personalized, day-by-day itineraries in seconds. Inspired by AI assistants like ChatGPT and smart travel apps, I wanted a tool that makes travel planning seamless, fun, and budget-conscious.

What it does

The AI Travel Planner generates complete travel itineraries tailored to your preferences and budget. Users enter:

  • Destination city (e.g., Tokyo, Paris)
  • Travel dates (up to 14 days)
  • Total budget
  • Optional preferences (e.g., vegetarian meals, luxury hotels, must-visit landmarks)

The system produces a day-by-day plan including:

  • Hotels
  • Breakfast, lunch, dinner
  • Sightseeing spots with Google Maps links
  • Budget tracking per day
  • AI adjustments based on preferences

The app works with OpenAI or GitHub Models and is ready for DigitalOcean deployment, demonstrating production-level AI infrastructure.

How we built it

  • Backend: FastAPI (Python) for REST APIs and request validation
  • Frontend: Single-page HTML/CSS/JS app for responsive UI and itinerary display
  • LLM Service: OpenAI GPT-4o-mini or GitHub-compatible endpoint to generate itineraries
  • Infrastructure: Dockerized for deployment on DigitalOcean App Platform, optional GPU for faster inference
  • Data Handling: Budget-aware calculations, Google Maps link integration, JSON output for frontend rendering
  • DevOps: Logging, health checks, environment variable validation, and production-ready containerization

The AI prompts were engineered to produce realistic hotel, restaurant, and sightseeing options while respecting user budgets and preferences.

Challenges we ran into

  1. Balancing creativity and realism: LLM outputs sometimes suggested unavailable hotels or attractions. We implemented response parsing and validation to filter realistic options.
  2. Budget management: Generating itineraries within strict budgets required dynamic scaling of hotel and meal options per day.
  3. Deployment on DigitalOcean: Ensuring a smooth Docker deployment with environmental variables, API keys, and LLM access was challenging but rewarding.
  4. Input validation: Preventing invalid trips (e.g., negative budgets, trips longer than 14 days) required frontend and backend checks.

Accomplishments that we're proud of

  • A fully functional AI travel planner that generates day-by-day itineraries in seconds
  • Budget-awareness integrated into every suggestion, including meals and hotels
  • Deployment-ready container compatible with DigitalOcean App Platform
  • Flexible LLM provider support — works with OpenAI, GitHub Models, or any compatible endpoint
  • Clear Devpost-ready README with setup instructions, demo data, and screenshots

What we learned

  • How to engineer prompts for complex multi-step AI tasks
  • Building robust backend services that handle validation, errors, and external API calls
  • Dockerizing full-stack AI applications for cloud deployment
  • Importance of UI/UX for AI apps — users expect readable, actionable itineraries
  • Handling budget calculations and preferences dynamically in AI-generated outputs

What's next for AI Travel Planner — Smart Budget-Aware Itinerary Generator

  • Interactive AI travel assistant: Users can modify the itinerary in real-time (“add a museum” or “upgrade hotels”)
  • Multi-city trips: Automatically plan trips covering multiple destinations
  • Local events integration: Include concerts, exhibitions, and seasonal attractions in itineraries
  • Mobile-friendly redesign: For travelers on-the-go
  • Integration with DigitalOcean Gradient AI agents: Fully utilize cloud GPU infrastructure for faster AI inference

Built with

  • Languages & Frameworks: Python 3.11+, FastAPI, HTML/CSS/JS, Pydantic
  • Cloud & Deployment: Docker, DigitalOcean App Platform
  • LLMs & APIs: OpenAI GPT-4o-mini, GitHub Models (free), Google Maps links
  • Tools: Uvicorn, Gunicorn, python-dotenv

Demo & Code

Built With

  • digitalocean
  • docker
  • fastapi
  • github-models-(free)
  • google
  • gunicorn
  • html/css/js
  • openai-gpt-4o-mini
  • pydantic
  • python-3.11+
  • python-dotenv
  • uvicorn
Share this project:

Updates