IceBreaker

Inspiration

As we were thinking of projects, we instantly gravitated towards the humanities and its idea of mental health. As an area often stigmatized and overlooked, we wanted to find a way to counteract such obstacles that have made it so hard to confront.

The easiest solution to this was accessibility. Both in accessing support and who we want to support us

This took the form of features like anonymous communication--an ability to talk to people going through similar issues in a very low-stakes way. We also extend this feature for non-emergency scenarios by using algorithms that analyze photos or their social media or discord account to connect them with others of similar interests as another way of company

We also implemented a chat bot that can ask the user check-in questions to track their well-being and give treatment options if there is a concerning trend. What makes this feature facilitate communication even further, though, is that if a user does not check in for a certain time, the interface automatically notifies emergency contacts that we have the user prepare when first entering them on the app.

This was why we were so efficient and deliberate the past 24 hours. In any case, it goes without saying that each person's subjectivity makes a problem like this nearly impossible to solve within a hackathon—even a whole year's worth of hackathons. Icebreaker is here to provide at least a path that encourages empathy and understanding in progressing someone's journey toward mental health—a journey that we must make easier for those suffering one way at a time.

What it does

Build mental health profile for user
Generate daily check-in questions based on profile
Analyze interests for user's social media profiles (e.g. Discord), to help match other people with common interest

How we built it

zWe worked on drawing diagrams to understanding the problem by drawing the features we wanted and what implementations we required and the research we would need as data to "teach" the Ai about mental health. From there one person did backend, one did frontend, and the other did further research to better make the program be knowledgeable of mental health

Challenges we ran into

With so many AI driven features, the clear challenge was indeed the learning curve with understanding Machine Learning. All of us are in our second year and have minimal experience with extending applications to support AI, let alone doing it in a hackathon environment. However, we overcame this by being methodical and taking time in our creation process. We first listed potential projects, drew diagrams that supported our desired project, and finally streamlined the coding process with AI by having two people tackle understanding and implementing API and the other person doing independent research of scholarly and internet articles to be generated as further data in our machine learning features.

Accomplishments that we're proud of

It's a live demo that you can use

What we learned

Don't over-complicate it We learned the reaches AI can have in many applications including mental health evidently. A lot of the times people worry about the ethics of mental health but we hope our project proves with enough effort you can make AI even empathetic enough to tackle mental health