CompostED

AI Chatbot that answers questions regarding composting and its benefits!
Compost Simulator!
Home page!
Our team's page and details!

Inspiration

Eugene Hwang's passion for composting was a heavy inspiration for the creation of CompostED. Branching off of his existing non-profit CompostTogether, Eugene believed that his initiative would be excellent to translate onto a program that teaches the importance of composting and the benefits that come with it, allowing us to create a more healthy and sustainable world.

What it does

CompostED has a composting simulator and a chatbot that is fine-tuned to answer questions related to composting. The composting simulator encourages players to learn about the scenarios in composting and the importance of greens and browns. The chatbot, Compostbot, answers questions that players may have regarding composting and its benefits.

How we built it

CompostED is built with Python and libraries. Our frontend was built with streamlit and some dabs of HTML and CSS to make it look pretty. Our backend is pure Python. Eugene and Cheech split their roles among the different parts of the project. Eugene took lead into creating the simulator whereas Cheech had the responsibility of creating the AI chatbot. In the end, they together worked on the about and home page.

Challenges we ran into

Early into the hackathon, Eugene and Cheech had struggled to come with a concept. It was not until approximately 5 P.M. That Eugene and Cheech had made significant work into their project. However, even more time was spent learning how to use Github and how to properly make push and pull requests. A major challenge was in regards of pushing and pulling. The first problem arrived when Github would not accept over 100 MB of a file, that was pyTorch. The other problems involving Github was with applying changes that were not in reference. Another problem that arrived was with images and styling them in Streamlit, in the end, making the images large solved the issue in terms of centering. Another challenge was with the version of sqlite3. No matter what, it would cause trouble with chromadb when we would run our application on streamlit. Turns out, it was just 3 lines of code that fixed the issue after hours of thought. Lastly of course, sleeping enough was a challenge on a different level.

Accomplishments that we're proud of

We are most proud of the project that we have managed to build within the time given to us. CompostED was an idea that was created late afternoon and finished late into the night. The little celebrations we had after solving major bugs in our code were some of the greatest coding reactions ever. We are also proud of the fact our chatbot works properly. Lastly, we are proud we had created such a creative project as first time hackers!

What we learned

We learned that we still have a lot to learn about concepts in how to be a developer. What learned most at hackNYU was how to use Github on our own. We had learned how to collaborate properly on a project with the use of pushing and pulling our work so we can work effectively. We learned how to use streamlit as the frontend of our project, in the past we have worked with flask. Lastly, we have learned what it's like to participate in a hackathon, especially in such a large one like hackNYU.

What's next for CompostED

Improving the chatbot with further data to fine-tune it. Making the simulator far more realistic with the factor of worms and other promoters of composting. We look forward to how others will use CompostED and their suggestions to our project.