Audio to text | Post-processing

[chenlung]

Tero Subtitler uses the models of Whisper for its audio-to-text feature, but the results are typically unsuitable for direct use.

As part of a post-processing feature in which the generated script is refined, what options would users like to see?