Motivation

Ask yourself this: what kind of music do you want to listen to right now? Is it taking you a while to choose even one song? Probably! Think about this instead: do you want to listen to Lana Del Rey? Yes! Do you want to listen to Drake? Yes! How about Katy Perry? Nope! Trying to craft the perfect playlist in your mind is difficult, but binary decisions like these are quick and easy. That's the idea behind SwipeTunes.

Summary

SwipeTunes is a playlist creation webapp designed to help users create the perfect playlist for the moment in the shortest time possible. Simply choose a song or artist to start with, and SwipeTune will present you with relevant back-to-back music suggestions that you can quickly swipe right to add to a new custom playlist or swipe left to reject. In less than a minute's time, you will have created an hour long playlist filled only with songs that you want to listen to! SwipeTune is perfect for creating the perfect workout playlist or choosing what songs you want to listen to before your drive to work.

SwipeTunes was a Top 16 project at HackPrinceton Fall 2015.

What's Next?

Although SwipeTunes was conceived as a web application that could be used as standalone music player, there are other platforms where it would perform better. For example, SwipeTunes as an iOS/Android application makes sense because the primary use case for such a feature is for on-the-go playlist creation. However, few users would be willing to download a separate application just for making playlists. I believe that SwipeTunes would thrive as a feature on an existing music player such as Apple Music, Spotify, or SoundCloud, as it would allow users to continue using their preferred music service with the added benefit of quick and smart playlist generation. Also, it saves the work of trying to create a new music service in a highly competitive market.

I think my next step for SwipeTunes would to be create a YouTube-Chrome extension that would allow users to create SwipeTunes-style playlists within their browser. YouTube is basically a music service, as you can find almost any song on the site, and a lot of people are currently using it as one, so it's a great starting place for getting this idea into the hands of actual users. That is of course unless SoundCloud, Apple, or Spotify would like to hire me to build this for them!

Some challenges with using YouTube is determining whether or not a video is a song and should be included in search results, but this seems like an interesting machine learning problem and a good opportunity to learn more about ML. This idea can also be expanded to play music from a variety of existing music platforms from your browser such as SoundCloud and Spotify, and additionally combine and take advantage of their search algorithms to find the most relevant music suggestions as determined by all exisitng music services. I could look into developing my own search algorithm for determining the best related songs, but other companies are already dedicating a ton of resources to just this and it might be best to focus on other ideas unless I find the need.

Technology Used

HTML/CSS, Javascript, Spotify's REST API

Github:

Devpost: