terratalk

Your AI Urban Assistant. Find georeferenced objects with simple queries like 'bus stations near Bryant Park'. Our app uses NLP to turn your words into results, fast.

Comment

Inspiration

We've all been there—trying to find a nearby service or landmark but getting bogged down in endless keyword tweaks and filter adjustments. Our goal was to transform this experience, making it as simple as asking a friend.

What it does

Streamline your search for georeferenced objects with our program, which uses OpenAI's NLP APIs to understand queries like 'bus stations near Bryant Park in 5 minutes' without keywords.

How we built it

The backend works by using ChatGPT to create an Overpass query based on the user's input. It then retrieves and parses this query, sends it to the Overpass API, and finally displays the map and locations obtained from the response. MongoDB Atlas enhances speed and personalization by caching.

The frontend was built with React and CSS and incorporates a simple login systen that parses the database for the user's personalized information.

Challenges we ran into

We struggled with accurate latitude and longitude plotting on map. However, after reviewing documentation we were able to figure it out.

Accomplishments that we're proud of

Despite the fact that that prior to the Hackathon all of us hadn't touched some of the main technologies used such OpenAI's API or React, we were able to quickly learn them to a functional degree.

What we learned

What we learned

  • Integrating NLP with Maps: We learned to bridge natural language processing with geographic data, turning user questions into map queries.
  • Geospatial Challenges: Tackling the accuracy of map locations taught us about the complexities of geospatial data and the precision needed for reliable results.
  • Caching for Speed: Using MongoDB Atlas for caching opened our eyes to how backend optimizations can drastically improve user experience by speeding up data retrieval.
  • React Skills: Developing the frontend with React sharpened our skills in building dynamic user interfaces and managing app state efficiently.

What's next for terratalk

With the nature of our program, it makes sense to expand to a mobile app. With this development, accessing georeferenced information becomes effortless and immediate, enhancing the way individuals interact with their surroundings, especially in urban areas.

Built With

Share this project:

Updates