We are interested in doing what you have done for music but to finger mp3 from podcasts, I don't know Go, but I am interested in knowing what part would need to be adjusted to work with mp3 and mainly spoken voices as podcasts don't often have music?
This would allow us to protect clients work, by scanning for copyright material used by unauthorised users. - Kinda like what youtube does but for the open RSS community.