16 million Americans are depressed; at least 20,000 people commit suicide because of depression every year. Being left alone at the most vulnerable time is both scary and dangerous.

Like all medical problems, prevention is the best solution, backed up intervention. Given depression is a very private problem, our motivation was to design a simple to use and engaging app that helps our users understand they mental well-being, without feeling being monitored.

What it does

EmMe engages with our users through texting. We send our users 3 daily "check-ups" customised to their everyday activities. Their responses are fed into our system and are analysed using IBM's Natural Language Understanding system to identify their happiness, joy, anger, and sentiment. EmMe tracks the users mental well-being overtime and builds rapport with the users by responding to their messages and sharing with them their emotional highs and lows in an intuitive chart.

EmMe also understands that sometimes user engagement and beautiful graphs are not enough to protect the most vulnerable and alone. To keep our users safe from self harm, EmMe uses carefully crafted nudging responses to steer the users to get help and when absolutely necessary, EmMe will reach out to the designated best friend on behalf of the user.

Challenges we ran into

Firebase and NLP were totally new to the team. Integration with the database - extracting information was the most difficult part to get working. We are very pleased to have been able implement these new tools to bring life and purpose to our project!

What's next for EmMe

EmMe will continue to improve with machine learning. Texting is the least resistive way to engage with a vulnerable user. As our knowledge of the users grow, we will be able to implement additional channels to help our users, from voice to video recordings as well as direct sync from calendar and geo-location to make our “check ups” more relevant and personable.

The end goal is to give EmMe the emotional intelligence to be the most trusted emotional companion.

Built With

Share this project: