The number of people who struggle with mental health is greater then ever, and is continuing to grow. We both personally know our closest friends who struggle with mental health issues, and we want to know what the best thing is that we can do for them. Developing a platform where people can express how they like to be treated during times of high stress, as well as be notified about what your potential mental health cycle could be, is what Espect solves.
What it does
Espect monitors one's day to day mental health level. This is done by inputting 6 different categories of mental health influences, and then uses machine learning to determine what ones mental state would look like in the near future. It is also used as a platform for people to share what is best for them, in terms of what other people can do, in order to ensure the best treatment and resources.
How we built it
We used Microsofts Azure to develop the machine learning algorithm, and used Apple's Xcode for the app interface. Stop by the demo or check out the GitHub to learn more.
Challenges we ran into
Some challenges we ran into were getting the back-end (Python, Azure) to connect with the front-end (Swift, iOS). The demo will be done using 2 machines due to this. Some other challenges we overcame were having to pivot and change directions completely about 6 hours in to the hackathon. Plus, all of the languages we worked with, we had no prior experience with, but it was a fun learning experience.
Accomplishments that we're proud of
Using Illustrator, Azure and machine learning was really cool. We also learned Swift and Python over the course of this challenge. We're very happy with the way this turned out.
What we learned
We learned how to use software such as Illustrator and Azure as well as app development and implementation of machine learning algorithms. We also learned about time management at a hackathon and how to have good communication between front-end and back-end devs.
What's next for ESPECT
We hope we can get the app integration working and hopefully get an Android version working as well. As for features, we hope to incorporate passive data such as weather, news and local politics into our inputs, as well as things such as Fitbit/smart watch data. This will help diversify our inputs in case the user forgets/is unable/doesn't want to enter data for the day.
Come chat with us if you have any questions :)