sometimes spotify just doesnt cut it. We named it Osusu after the Japanese word for recommendation (osusume).

What it does

Recommends songs and artists to create a user personalized playlist based off of user history.

How we built it

Blood, sweat, and tears. We used the Million Song database to cross reference attributes of music that a user already "likes" to suggest them other music. We used Python and SQLite.

Challenges we ran into

We were initially going to use Java, but we ran into some issues with outdated packages, so we switched to Python (also better to work with SQLite). Also, lack of sleep.

Accomplishments that we're proud of

Learning how to use the database. Also, I was getting cold feet and was anxious about my lack of experience and was thinking about jipping. Glad I showed up!

What we learned

Learning how to use the database. That Python 2 is discontinued. How to use SQLite #swag.

What's next for Music Profiler

Being named Osusu instead of Music Profiler.

Built With

Share this project: