Sample from Ignite 2017 talk, that uses the Speech API to transcribe audio files in a document library - View it on GitHub