Share the post "AI-based music startup Musiio helps major music distributors to process and recommend the music!"
Not so long ago, a Singaporean startup “Musiio” was sold to the Sound Cloud for presumably 10$ million, which is a worth of the company at the beginning of 2022. SoundCloud is a very popular and famous music distribution platform where every single person can upload their tracks, beats, receive listeners and subscribers. There are a lot of famous artists that used SoundCloud to gain their first fan base before becoming globally recognized such as Denzel Curry, Suicide Boys, Night Lovell, etc. But, without a proper system that would process the tracks uploaded, assign specific metadata to them and put in a specific playlist or list, tracks of the artists mentioned could be simply lost in a flow of 300 million sounds that are present on the platform and would never find their listeners or interested parties. Recently the flow of new tracks from the young artists increased significantly and SoundCloud started struggling with structuring and recommending the matching tracks and playlists to the listeners, that is why SoundCloud showed a big interest in the technology that Music-Tech startup Musiio developed and as a result decided to acquire the startup for further development and integration.
Musiio is a Singaporean based startup that was established in 2018 by Musician Hazel Savage and ML Engineer Aron Peterson. Musiio uses Machine Learning and Deep Learning to identify the unique parameters of the track tag it with a set of tags such as BPM, Genre, Mood, Energy and record those parameters in the track’s metadata for further predictions, structuring and recommendations. For now, startup possess 3 ready to use APIs such as Music tagging, Music search and Music recommendation. Music search allows to search tracks similar to the track inputted by making use of tags and tags matchings. Music recommendation allows to learn the preferences of the listener by continuously processing tracks listened and as a result creating a personalized playlist for him. Here is the demonstration of the high-level pipeline that Musiio’s implemented for their algorithm:
In 2018 Musiio cooperated with a nonprofit organization called FMA (Free Music Archive) that possessed a database of over 120,000 tracks from different non-profit creators. It was the first major project of Musiio that proved the efficiency of their algorithms. According to the head of FMA, Musiio was able to attract thousands of listeners to the tracks that were “rot” in the FMA database for over 8-10 years. After success with FMA, Musiio started actively cooperating with different globally known companies such as Vans, Auribus, BeatStars and successfully implemented its solution in their recommendation use cases.
After 2 successful years of operation, Musiio became a part of SoundCloud ecosystem with the right to operate as B2B company. SoundCloud claims that Musiio’s tagging algorithms will significantly help to search the tracks that match the interests of the listeners, handle and structure the flow of tracks that’s being uploaded to the platform daily and of course to create personalized playlists and music selections.
References:
- https://techcrunch.com/2022/05/03/soundcloud-acquires-musiio-ai-music-discovery/
- https://www.theverge.com/2022/5/4/23056806/soundcloud-acquisition-musiio-artificial-intelligence-ai-hits
- https://techcrunch.com/2018/04/20/musiio/
- https://musiio.com/?utm_source=blog&utm_medium=link&utm_campaign=blog_redirect