I’m Sergiu, the author, and maintainer of SoundFingerprinting open source library. Over the last 8 years, SoundFingerprinting has grown to a fully functional framework for audio search. During these years, I’ve developed two open source storages that enable storing audio fingerprints. Both are available under MIT licenses and serve the purpose for simple use-cases where you have up to 1000-5000 tracks you would like to store and retrieve later.
As the framework matured, I’ve been asked multiple times to provide more scalable storage which could store tens of thousands of tracks, but most importantly, storage that will provide fast lookup times.
That is how Emy was born. It is a storage that has been battle-tested in production, monitoring hundreds of radio and TV streams. If you have vast amounts of audio data you would like to search for specific tracks or ads, Emy is for you.