We wanted to provide a simple way for people suffering from mental health issues to tell other people about it. We know it is difficult to tell people around you when you are suffering from depression or anxiety, and we wanted to make that process easier by (1) quantifying the disease through health metrics like sleep, and (2) providing an automated tool to let the necessary parties know if symptoms are recurring.
What it does
The app ingests user data from the fitbit API for each registered user, then runs an anomaly detection algorithm to see if sleep patterns over the previous 5 days have been significantly different than normal. If anomalous behavior is detected, the program asks for permission to send a text message to all members of the patient support network, informing them that the patient is suffering from symptoms again.
How we built it
We built our web frontend in React/JS and built out most of the algorithm detection algorithm in JS as well. The Fitbit API provided the user data after OAuth authentication, and used the Twilio.js library to send messages. We planned to develop a user data-storage system via the built-in NoSQL database in Firebase, but didn't have time to implement this functionality.