Inspiration

We researched ideas on how to reduce waste effectively and realized it likely starts with the consumer. Using this, we decided to increase awareness of these alternatives directly to the consumers with a browser extension.

What it does

Our Chrome extension overlays a pop-up when the user is on a product page on Amazon.com. Using a Google Gemini API, the AI identifies keywords in the URL and categorizes them by departments on Amazon.com. Using the identified keywords, it produces search results that are more sustainable and allows the user to click on a product, being redirected to its corresponding product page.

How we built it

We utilized JavaScript (.js), JavaScript Object Notation (.json), Cascading Style Sheets (.css), and HyperText Markup Language (.html). We utilized the Visual Studio Code IDE to house all of our programming and extended the workspace to Google extensions to display the finished project.

Challenges we ran into

  1. Given the complexity of our idea, we had to learn and understand the syntax that our project utilized.
  2. Initially, the AI was picking up the wrong keywords, which altered the results that our extension presented.
  3. Once we got the AI to produce results in the same category as the item the user was viewing, we had to tune the accuracy of the results. We ensured that the AI would include sustainable words in its search results, ensuring that the recommended products were actually valuable to the user. ## Accomplishments that we're proud of Obviously, with the time constraints this challenge presented, we are proud of the current state of the project. Compared to our previous projects, we took a step further in implementing an AI API, which proved to lighten our workload and was still a great learning opportunity. ## What we learned We learned how to utilize and configure the Google Gemini API ## What's next for SustainSearch Our next step for SustainSearch is to expand the compatibility of websites so that the extension works on more e-commerce websites. Additionally, we would like to improve the accuracy of the results the AI produces, as we found that occasionally, it can make mistakes.
Share this project:

Updates