-
main page
Our inspiration
It was us hanging out and having a specific mood and not knowing what songs best fit that mood.
Input
The project takes input from the user, and based on the description, the project will output specific songs based on OpenAI's and spotify's recommendation engine.
How we built it
We built it by having a front end for the project that takes the text input from the user, and then it updates with the specific songs from the back end. The back end basically gives the input to GPT 3.5, and the AI returns specific levels for music characteristics such as danceability, popularity, etc. Then the project takes those values and inputs them into the Spotify Recommendation API to recommend songs for the front end.
Challenges we ran Into
We were passing in the same responses into the APIs and making sure no data got lost.
Accomplishments that we're proud
Being able to make the specific prompt for GPT 3.5 and converting them to a format that the spotify API could utilize.
What we learned?
We learned the specifics about API calls and how to utilize them in a full stack application.
What's next?
Being able to generate spotify playlists with the music that is recommended would be next.
Built With
- next.js
- openai
- spotify
- tailwind
Log in or sign up for Devpost to join the conversation.