Cocodrilo

• 

  Homepage of Cocodrilo

Inspiration

Our inspiration was to create a program called Cocodrilo that helps with sustainability while incorporating a generative AI model that can dynamically interact with the user.

What it does

The chat bot is designed to ask the user a series of questions about their lifestyle to help the bot understand the general carbon emissions that the user is responsible for. The bot will give the user an estimate of their carbon footprint and provide them with some ideas to help decrease their carbon footprint.

How we built it

We built this chat bot using GitHub Copilot to help us have a starting website with some basic UI elements. We then refined the UI to look for user friendly. We spent some time training our AI model with Google AI Studio, which allowed us to refine our prompt to ensure that the bot was behaving as intended. We wanted a bot that did not provide a verbose response, and a bot that provided relevant questions that the user can easily relate to. Once we set up the bot and UI, we hosted our server using Azure, and we created a domain using namecheap.

Challenges we ran into

We ran into issues with our initial setup because we were not familiar with how to set up Google Gemini to work on our server. We were able to fix this issue by reading the documentation on the Google AI website. We also ran into issues related to the training of the AI model because the prompt building had to be super specific.

Accomplishments that we're proud of

We are all participating in our first hackathon, so we are very proud that we were able to create a working product with a clear workflow. We are proud that we were able to create this program within the time period that we had.

What we learned

We learned a lot through participating in this project. We learned how to use copilot to help us create a good starting point for us to create our program more efficiently. We also learned a lot about using Google AI studio and Gemini in our program by reading all the documentation. We also learned a lot about continuous integration using Azure and GitHub.

What's next for Cocodrilo

We would like to implement hardware that works with Cocodrilo to help create more accurate carbon emission tracking. We would like to create something that attaches to a car so that you can track the carbon footprint of a user based on their actual driving metrics.

Built With

• azure
• copilot
• gemini
• github
• html
• javascript
• namecheap
• node.js