do you handle things like dialectic differentiation or entity detection automatically, or is that something left to the user once the transcription is done? | discoverkit | discoverkit