When you first log on to the app, you’ll let Fynda know what genres of music you listen to (you can pick as many as you like. Seriously). Fynda will then serve you up song recommendations based off what’s popular in these genres. Then you need to get swiping! The more you swipe, they better your song recommendations will be. Think of your swiping actions as a constant feedback loop for our AI. No feedback, no change.
We want this app to be a constant experience of music discovery, so we made sure you can find (“fynd”) our top selections for you, featured artists and songs, profiles with similar music tastes, and local music events all in one place (the Fynda page). If there’s a song you know you love, you can also search our library and make sure that song is factored in to your recommendations.
Our goal is to make sure you love every song recommendation you receive. Fynda constantly is updating your “music imprint,” and we designed this to be a hands-off experience. But if you think there’s a song in your imprint that shouldn’t be there, or your music tastes evolve, or you just want some different recommendations, you can always manually edit your imprint so you’re getting the best recommendations possible.
Fun, addictive swiping actions get you excited to discover new music (not that you really needed the extra motivation)
We use machine learning and artificial intelligence to bring you the best song recommendations possible
Our AI automagically creates collections for you based on mood and genre, which can be exported to your favorite streaming service
Not quite sure what kind of music you listen to? Gain clarity with our music imprint, which breaks down your listening habits by genre
We're not saying music is a predictor of good friendships ... but music is a predictor of good friendships
Take those headphones off and discover local artists. We know that good live music is a transformative experience.
Using a popular wave form analysis framework by Panotti, we combine TensorFlow deep learning convolution neural networks (CNN) to study audio snippets as WAV files after they’re converted and proportionately framed in JPG spectograms, to derive a unique matching model for every user’s swiping collection.
This provides a unique experience that is personalized for every user on our platform allowing for an evolving level of specificity for recommending new songs and matching all your favorite music styles with other users.
We are easy to reach! Email the our team directly, contact us via phone, or visit us in Denver, Colorado!