Hong Kong Travel AI Assistant

Provides personalized recommendations, trip planning, and real-time information about weather and traffic conditions.

Comment

Inspiration

We wanted to make exploring Hong Kong easier and more fun. The city has endless sights, changing weather, and busy transport. Travelers often struggle to plan day trips that match their interests, weather conditions, and traffic. We saw an opportunity to use AI and real-time data to solve that problem.

What it does

Our AI assistant can give personalized travel advice in real time. It suggests attractions based on your interests, the day’s weather, and traffic conditions. You can build and edit itineraries, view transit updates, and filter spots by theme or activity. Everything updates automatically so you always have the best plan.

How we built it

We used the Amazon Q Developer to help us promote the code and fix bugs. We used a three-layer architecture:

  1. Backend in Python with FastAPI for AI recommendations, weather, and transport APIs
  2. Node.js + Express as a proxy for caching, aggregation, and error handling
  3. React with TypeScript on the frontend for interactive maps, responsive design, and animations
    Data lives in PostgreSQL with Redis caching. Docker and AWS manage deployment, and the Qweens 3 14B model powers our AI logic.

Challenges we ran into

  1. Integrating multiple real-time APIs and keeping them in sync
  2. Designing AI recommendation logic that balances user preferences, weather, and traffic
  3. Optimizing itinerary generation so it runs quickly under load
  4. Ensuring the UI stays smooth with frequent data updates
  5. The token problem when calling the api key too many times
  6. Do not know how to use Docker
  7. Many Many Many Bugs!!!

What we learned

  1. How to manage and cache real-time data from different sources
  2. Best practices for combining AI services with traditional APIs
  3. Techniques for prioritizing and scheduling activities under changing conditions
  4. How to architect a full-stack app that remains maintainable and salable
  5. The important of teamwork

What’s next for Hong Kong Travel AI Assistant

In the future, we can add more local event and festival data for time-sensitive recommendations. Also, expand language support and accessibility features is very important to attract more users. Moreover, partner with local guides, restaurants, and attractions for exclusive offers is also a good idea for us to provide more detail services.

Share this project:

Updates