cjpais
Feature Request: Add option to reprocess audio transcription in History
When recording voice notes in Handy, the transcription sometimes does not capture the full content of the audio. For example, a 17-second recording may only be transcribed into a few words (see attached screenshot).
It would be very useful to have a “Reprocess” button in the History view, allowing users to send the same audio file back through the transcription module for a second attempt. This would give the system another chance to generate a more accurate transcript without requiring the user to re-record the message.
Proposed solution: • Add a “Reprocess transcription” button next to each recording in the History view. • On click, the audio file is resubmitted to the transcription backend. • Replace or append the new transcript, possibly showing versions if multiple attempts exist.
Benefits: • Improves user confidence in transcription results. • Saves time compared to re-recording audio. • Provides flexibility for cases where transcription accuracy fails on the first pass.
Example: • A 17-second voice note was only transcribed as “Petite question de travail.” • A reprocess option could potentially return a more complete transcript closer to the actual spoken audio.
Yeah I would like this as well. Also need it for debugging. I'll add it when I get a chance
