What inspired us?

Our team acknowledges that the world has been locked down for the past couple months, missing some of the most pivotal events of the year, and we believed that during these times, many small businesses are worried about when to open, how to stay calm despite hectic circumstances if they shouldn't open, etc. Acknowledging this problem, our team came up with a solution, where we take in a user's (business) location , and based on the location, find out various Covid-19 related characteristics about the area such as total deaths, total infected, total hospitalized among others, use those characteristics for our machine learning model, and let the user know whether or not it is a good idea to be reopening their business right now. If so, what precautions and guidelines should they be abiding by in the store, and if not, what should they be doing at home to keep themselves safe as well as be prepared for when it comes time for their business to open.

What it does

Breakout COVID can be used as an app and website, which uses Machine Learning model to suggest business companies whether or not they should close down their business or not. This issue started to appear earlier last month, where companies, including Apple, started to close down their stores due to the Coronavirus outbreak. This caused many other companies to debate whether it is too early to or too late to close their business. We realized that based on the data we have about COVID-19, we can find solution to this by training machine learning model to exactly this. The mobile app, which is developed with flutter, and webapp use this machine learning model to tell its customers whether they its too risky or not to close their business. Along with this prediction, the users are also able to look at a bar chart to see the COVID-19 related data with ease and all in one place.

How we built it

Our team developed both a mobile and web application for our project, web being made from flask and mobile being made through flutter. For the flutter application we coded in Dart and used different packages such as Flutter Chart Packages and Firebase for user authentication.

Challenges we ran into

Although CovidHacks project submissions only lasted about a day, we aimed to integrate the most ambitious of our ideas into the app, and definitely they forced us to overcome challenges we faced. When eagerly scouring the internet for resources to solve the problems we were encountering, we would just wonder on end about how to do it, but as soon as something clicked, no matter big or small, we would get enormously excited and yet another burst of motivation would come along. Definitely one of the more difficult challenges our team came across was connecting our app to an unsupervised Machine Learning Model which then outputs a decision as to whether or not based on information in the area if it is safe to be open.


Creating and Hosting Machine Learning model and creating a user friendly website and app to help businesses make decisions whether they should close or not. Firebase Authentication and Firebase Database to authenticate user and store user data.

What we learned?

We definitely learned from this experience that nothing is as simple as it seems. When going through a lot of the components which were part of our project, ones which we believed would take 30 mins - 45 minn and wouldn't be mind boggling ended up taking hours on end. On the flipside, this was a life lesson, but we also learned a lot more about the languages / frameworks we were working in. Having not worked in flutter too many times for too long, it was a great experience to adapt and find solutions to build a complex app all in a day and I feel like this is the best kind of environment to learn and grow in.

What’s next for Breakout Covid

We plan on adding more functionality for this app by improving the machine learning model, making an api for the website and app, and adding more features that would be useful to users. Improving the UI is also a goal that we want to achieve.

Share this project: