From a technical standpoint, I'm trying to do exactly that. We have a basic standalone watchOS app that, when you hold down the button in the app, takes the audio from your voice and streams it to a server for transcription and further processing. Unfortunately we've been blocked by the inability to use websockets for this implementation and I'm looking for alternatives. Since we're trying to stream audio in this manner, how can we get access to the low-level networking functionality?