The social distancing measures necessary to prevent further spread of the Coronavirus have led to extreme feelings of isolation and loneliness for many at a time when they need human connection more than ever. These feelings combined with the uncertainty and fear that many are experiencing has led to the Coronavirus outbreak being very difficult mentally and emotionally, particularly for those with underlying mental health conditions. We became inspired to work on this project because we believe nobody should have to feel like they are going through this difficult time alone.
To learn more about how this newfound isolation has been affecting people we spent the first day of the Hackathon conducting over 30 interviews of people from a variety of different backgrounds and age groups including an elderly woman isolated in her home, a college student taking classes virtually, and a restaurant worker who lost her job. While they each had different stories to tell, they all expressed to us how feelings of loneliness had been taking a toll on them and how they were desperate for new outlets to meet and connect with people while staying in their homes.
Additionally many people seemed to be not doing much while at home and were struggling to find motivation to stay productive yet had ambitions such as starting a company or learning something new. We wanted to build something that would turn this time of hardship into one of innovation, learning, and growth for people by helping them find others to keep them motivated to be productive or to capitalize on their free time to fulfill an ambition.
What it does
After hearing all of these stores, we decided to build Quarantine Buddy. Users sign up for the platform, filling out basic information about themselves such as their age, where they’re from, and what they’ve been up to these past couple of weeks. Additionally, they can specify what they are looking for in a Quarantine Buddy, whether it’s someone to do Zoom yoga with, to collaborate on song ideas, to start a company with, or just to meet someone new. Our optimal matching algorithm then pairs and introduces people who are good fits for one another or who are looking to do similar things with their quarantine buddies and schedules times for them to connect and meet via Zoom.
How we built it
To complete the project in the given time frame we had to effectively prioritize and mitigate different tasks. We decided that we should try to get a website up and running as quickly as possible so that we’d actually be able to have real users sign up and get connected with one another over the weekend. This led to the decision to create the build the current quarantine buddy website with its simple form to fill out as a Minimum Viable Product using Wix before going forward with creating a robust and scalable tech stack.
We built our optimal matching algorithm using a random forest model. It first cleans and modifies the data that the user puts in, then weighs it using designated weights for each category and creates a similarity ratio using term frequency (TF-IDF) of the vectorized object. Each person is then paired with the person who has the highest similarity ratio and the results are exported to a CSV.
Challenges we ran into
A challenge we ran into was designing the user interface of the website to be used by people of all age groups. After speaking with both younger and older audiences, we came to learn that a website that seemed to be targeting older audiences would turn off the younger audiences to signing up more than the other way around. This led to the decision for the time being to design the website itself to appeal visually to millennials (despite the fact that the product itself is to be used by all age groups).
Accomplishments that we're proud of
Since launching the website three days ago over 400 people have signed up for Quarantine Buddy of ages 18-70+ from places ranging from New York to Australia to South Africa.
We have already heard amazing stories of people and their Quarantine Buddy meeting up via Zoom and doing wonderful things ranging from writing songs together, discussing plans to start companies, and making Tik Toks. Some other matches include educators from different universities who discussed their switches to online learning, people from different generations and talking about their experiences, and mothers from different parts of the world discussing how to help their children while they go to school virtually.
What we learned
We have learned so much just from the matchings we have done so far. For one thing, while we found that people are very eager to connect with others (especially during this time), that people are inherently bad at scheduling times to meet with one another themselves. They are much more likely to hop on a Zoom call when the time is chosen for them rather than going back and forth trying to figure out scheduling and logistics. This taught us that we need to include scheduled times to meet when we introduce the quarantine buddies to one another.
Additionally, we learned that some of the best matches came from differences, not similarities. People of different age groups and backgrounds tended to have some of the best conversations and ended up being some of the best pairings. Furthermore, we found that people were able to feed off of each other's motivation. For example, if we matched two people taking online classes, they were able to motivate each other to stay productive even though they were not taking the same class. Buddies were also able to encourage each other to finally get that project done or to start writing that screenplay.
This taught us that going forward we should consider iterating upon our current matching algorithm to account for the fact that similar people may not be the most optimal match.
What's next for Quarantine Buddy
From all the feedback we’ve received and will continue to receive from connecting people on the current site, we will be able to further perfect the concept and to ensure that we are tackling this problem as best we can before going into full product development mode in which we will use professional and sustainable development techniques to build out a more robust version of the product. .
Additionally, while we are now focused on connecting people during the Coronavirus outbreak we see many applications for the matching algorithm and platform we’ve built such universities being able to connect their students learning remotely, companies ensuring their employees are engaged and collaborating while remote, in addition to hospitals introducing patients who have similar medical conditions.
Social introductions have been rapidly shifting online over the past couple of years, as evident by online dating and meetup apps, and the possibilities for connecting people, not only at this difficult time, but going forward are endless.