Finding people with common interests was hassle during the pandemic and this was common in our team. We found a niche in this problem and therefore wanted to build a solution for it.

What it does

Stuck at home attending stupid online classes 👨‍🎓? Got no Among us community and wasting hours on public servers 🎮? Wanting to discuss your Linear Algebra homework at the last minute 📖? On a quest to find the next perfect Netflix series to binge watch and discuss with a buddy 👀😘? The pandemic sucks, we know it - so we built 🛠 GGBuddy - the perfect place to find a GGBuddy to do anything and everything. Our advanced machine learning algorithms find you the perfect match to make 2020 great 🚀 again!

How we built it

To build it, we used HTML, CSS, JS, Python, Mongodb and Node.js. The website was built with HTML, Bootstrap, CSS and Javascript. After that, we made a form using Mongodb and Node.js and later with third party software. After connecting from different users, we trained our machine learning model using tensorflow on 500+ entries and tested on 300+ entries. Although, we got an accuracy of 89.95%.

Challenges we ran into

We faced various challenges during the hackathon. First of all, working in different time zones, then aligning ourselves on a common goal. The trial and error method of different machine learning models like K-n nearest neighbours, linear regression, and finally tensor flow was a tedious task.

Accomplishments that we're proud of

1) Completing the MVP of the project 2) Meeting new people 3) In the process of the hackathon, we learned Mongodb and Node.js. 4) Working as a team

What we learned

1) Mongodb 2) Node.js 3) And most importantly: teamwork

What's next for GGBuddy

This is definitely not the end for GGbuddy. We see it as a great idea because it focuses on solving a problem which student's have been facing for a long time. We will probably find a developer to improve the User interface. We will give a thought on the business model and improve the machine learning model by training it on more data.

Share this project: