Mood on Repeat
Hosted: Link
Github: Repo
Goal
This project was completed as a part of Hack the North 2020++. The goal was to interpret a user's mood from their music choices and then suggest playlists to adjust their feelings.
From our Devpost:
Our project looks at your On Repeat Playlist on Spotify and interprets your mood. Then you can choose to find more music with a similar feeling or try some new playlists to change your mood.
I was responsible for the design and frontend development of the project. The goals for this situation were to work quickly from idea to implementation and create a simple UI to help any novice user connect to Spotify and see their results.
Process
After understanding our flow with my team, I designed the flow in Figma to have a plan while developing. Once the design was complete, I confirmed it with the team.
To develop quickly, I used a vanilla React app with Bootstrap. The only integration for this project was to link to Spotify's OAuth login.
The app was not deployed during the hackathon, but I later deployed the frontend on Netlify.
Next Steps & Lessons
Unfortunately, the backend for our project was not completed in time for the hackathon deadline. Going forward, I would not separate the team between frontend and backend so dramatically because I was not aware of the status of the backend throughout the process.
As the project was completed under a time crunch, there are sections of the UI that could be cleaned up. Additionally, the project was developed for the laptop screen size, it needs further work to make the mobile experience acceptable.
To complete the project, the backend would need to be completed, upon reflection I don't think that judging the user's mood based on their Spotify usage is the correct approach. My updated theory is that Spotify's current music recommendations probably serve this purpose already, though indirectly.
If I was going to approach this concept again, I would look at a data source outside of music habits to get the user's mood, for example, Samsung's watches offer a stress level measurement. I think an app which measures mood from a wearable, then suggests playlists, would be a more interesting product.