EcoLibrium

• 

  logo

• 

  post creation

• 

Inspiration

We were inspired by the TV screen above a garbage can directing people on where to put there garbage. We realized that even with the information about where to put our bottles, people still put them wherever they want. There was no incentive to recycle it. Thus, we decided to try to find an incentive to reuse garbage in an attempt to prevent them from ending up in the garbage.

What it does

EcoLibrium is a social media style app that allows people to post their arts and crafts projects that make use of old garbage. Similar to the 5-minuit crafts youtube channel. This should inspire people to reuse old garbage and provides an incentive to keep reusables out of the garbage.

How we built it

We built it using the python framework Flask. We divided our group of 4 people into front end, backend, and AI. For the first 12 or so hours, Roy worked on setting up the base functions for creating and viewing threads etc, Sudipta and Daniel worked on styling our front end and creating a Figma markup of what we wanted our website to look like, and Michael worked on creating our text summarization AI to be used in creating posts.

Challenges we ran into

Originally, our text summarization AI was supposed to be used in combination with a search algorithm to find and rank posts. However, between errors and lack of knowledge, we realized that it was way out of scope for what we were able to create in the 24 hour time period. Thus, we ended up just using it for generating descriptions for the posts. Additionally, we also ran into trouble creating a graph for our data as none of us had really ever done front end before. While Daniel had experience with Figma, neither of them had really done HTML and CSS. This made creating a graph especially difficult.

Accomplishments that we're proud of

We are proud of our team's persistence in getting through errors and sickness. While two of our members took a 'sick-leave' part way though, we were still able to finish on time. We were also able to build a functional text summarization AI within the 24 hour time period even though we had only ever used premade functions from over a week earlier. Lastly, we are also proud that we were able to create a social media style Post and Comment system without any prior experience.

What we learned

We learned a lot about API implementation especially from the Penseum presentation. While we ended up using huggingface instead, they were also a lot of help in troubleshooting errors and helping correct errors in our AI. We also learned a lot about front-end as we tried to create all of the styles from scratch without any experience.

What's next for EcoLibrium

We already have a domain set up, we just went able to launch the site within the time limit. We plan to launch the site in the upcoming week and iron out as many unknown remaining bugs as possible..

Built With

• colab
• css
• flask
• html
• javascript
• python
• sqlalchemy
• transformers