Dictation Transcription with IBM Watson or Google API

Now that I have transcription of voicemails by email working, I’d like to try my hand at using IBM WATSON in conjunction with the dictation module.

I did not see a post dictation script option as there is with voicemail to email or call recordings. If this exists, please let me know.

Is there a way mimic what the transcription module is doing with a custom context (I’m sure there is… any ideas on what this would look like would help).

OR, does anyone know where the dictation module saves wav files? I could run a batch script to look for files in this directory, then transcribe and email them if one person is using it (unless of course there is a subdirectory created for each extension).