We were motivated to build an app for social good that fit the theme of Pride. The idea behind Emotion Ally is that by understanding their emotions better, LGBT+ youth can better manage their symptoms, get support, and thrive like their peers.
What it does
It predicts the user emotions as per their thoughts. The user expresses what they're thinking and feeling by typing it. The program identifies how they feel and directs them to resources that could help.
How we built it
We designed with Figma. We built it using Flask as frontend and backend and used cohere for NLP which helps us guess the users emotions according to their thoughts.
Challenges we ran into
- Inexperience with the technologies: We had a mostly beginner team of 3. Most of us was using Flask for the first time and NLP was new to everyone. Two of us were also inexperienced with web development.
- Styling in Flask using CSS: Due the our inexperience, we struggled to match the application's design to the initial Figma design.
- Integrating different technologies: Integrating the NLP into the flask application was difficult because we struggled to communicate and merge them
Accomplishments that we're proud of
- Successfully brainstorm and agreeing on idea
- Using Figma for the first time
- We addressed our inexperience by diving into the new technology. One member learned Flask, and another learned NLP this weekend. Then we tried to combine our knowlegde in the app.
- We worked together across 3 timezones.
What we learned
- Project completion
- Design with Figma
What's next for Emotion Ally
- Facilitate user expression with speech recognition and/or face detection.
- Expand the range of emotions our model recognizes including more nuanced emotions like anxiety and depression.
- Securely storing user data so they can track their emotional progression over time.
- Flagging entries as signs of a deeper mental health issue and connecting the user with the help they need.
- Expanding our list of resources.
- Make the app more aestheically pleasing to invite more users.