STRINGR

TechCrunch Disrupt Hackathon Submission May 2015

Find Nearby Musical Collaborators

As a musician, finding collaborators for your musical projects can be difficult and time-consuming. We use tinder's validation mechanism to show you other artists' sound clips who live in your area that have been analyzed and sorted by sentiment score. By logging in via SoundCloud oAuth, your own clips are uploaded and scored by sentiment as well for other users to see. If you and another user upvote each other's tracks, you both receive a Twilio text notification and are able to chat with each other on the matches page.

Sentiment Analysis Scoring API We used SoundCloud API to get user_id for each artist and their tracks. The final score for each artist is calculated based on avg sentiment of each track, number of followers and the number of tracks uploaded. The json dump for each artist is periodically indexed using Elasticsearch. We then provide a REST API using Flask + ngrok.

stringrapp.herokuapp.com

Team

Brian Borge (brianborge)
Don Kahn (kahdojay)
Nick Mascia (nmmascia)
Salil Navgire (salilnavgire)

Built With

Submitted to

TechCrunch Disrupt NYC 2015
- Winner Disrupt Tickets

Created by

I worked on setting up oauth through Soundcloud, building the models, creating model tests, integrating circle ci.

Brian Borge
Worked on Soundcloud API to score artists based on avg sentiment on their track comments+other features. Indexed in Elasticsearch server and created a API using Flask(Python).

Salil Navgire
Data Scientist, ML and NLP hacker
Worked with the SoundCloud API to render tracks on the front-end using data from the machine learning algorithm developed by Salil.

Don J. Kahn
Product/Dev
Nicholas Mascia