Being a team of 3 college freshmen during the COVID-19 era makes it much more difficult to connect with peers and make new friends. We wanted to help make this process easier by recreating some of the smaller interactions that happen on campus during a normal semester that lead to some of the best friendships.

What it does

Discord is the main platform college freshman have been using to connect with each other socially and academically outside of class this year. More often than not, these Discord servers that students are a part of are massive with hundreds of people in them, and it can be quite daunting to talk in the server and connect with others.

As a solution to this, we created a Discord bot and a corresponding phone app. When a user has free time and is looking to talk to someone new, the user can tell the bot that they are "ready" to talk and the bot sorts through all the "ready" people and puts the most "compatible" people into a voice call as soon as it has made a match. This is that recreation of the feeling of "bumping" into someone.

After the call, the bot messages each member of the call individually and reminds them to go to the phone app and submit a survey on how much they liked the conversation and the people they talked to.

The bot then has access to all the survey submissions and can create and update "compatibility" between people. Then each time a user wants to talk, they are matched with not only people they've enjoyed talking to, but also people that the bot thinks they would enjoy talking to.

How I built it

We built the bot using python and the Discord API. The phone app is made using Glide.

Challenges I ran into

We had trouble getting the bot to always run on Google Cloud due to time constraints.

Accomplishments that I'm proud of

We are proud that we were able to submit a functional product!

What I learned

We learned that everything requires a lot more time than we anticipate.

What's next for BumpBot

We would like to expand to more platforms such as Slack. We would also like to improve the compatibility algorithm.

Built With

Share this project: