Rally

Inspiration

Inspired by the recent movements in the Cornell community, we wanted a platform where people could rally together for important social causes. In today's world, while we are so interconnected, it is hard to find a good platform on which someone can enact meaningful change. From community building to fundraising, no one platform covered everything we were looking for in our local communities. This is what inspired us to create Rally.

What it does

It's a platform for staying connected and building communities for good. It keeps people updated on current issues, helps organize meetups, centralizes community fundraising, and suggests local events that fit your interests. In order to keep Rally safe for everyone, we verify your Cornell email in order to sign up, but allow you to freely express yourself by choosing an alias within the application. Whether you are looking to get more involved in social change in your community or want to host the next big movement, Rally is the app for you!

How we built it

We built the web application on React.js using an Express.js backend and MongoDB as our database with Pinata being used for image storage for profiles. Additionally, we utilized a Python script with machine learning (BERT and XGBoost) to create an event classifier and recommender. The backend API for this script was built with Flask and accessed by our Express.js app.

Challenges we ran into

Coming together to develop so many different sections of an application simultaneously proved to be a difficult task. While we all had strong backgrounds in software development and machine learning, we had never participated in a hackathon prior to this event, meaning this was an entirely novel challenge. Overcoming these difficulties were not easy, but we are very proud of the app we developed in this short time.

Accomplishments that we're proud of

1. A sleek, modern UI.
2. Learning valuable new tools and skills on the fly.
3. Successfully implementing machine learning for event classification and recommendations.
4. Creating a fully functional app! ## What we learned
5. Frontend and backend communication.
6. MongoDB integration.
7. Using BERT for natural language processing.
8. The importance of Git for collaboration.
9. The Pinata API
10. Agile is hard to do half asleep ## What's next for Rally We’ll refine the backend and any design errors, including full-fledged email verification for students. We also plan to promote the platform after we leave!

Built With

• axios
• bert
• express.js
• flask
• javascript
• mlflow
• mongodb
• node.js
• pinata
• python
• react.js
• sckit-learn
• xgboost