This is a bad filler world that would be helpful if it was removed

Sometimes, filler words or hesitation add emphasis to a discussion - perhaps a new feature would be to not only edit the transcript text - but to also edit the same text in the audio (by keeping or removing the corresponding audio)

Since you now have the ability to detect different speakers in the same track, it would be nice to add the ability to split speakers to separate tracks for separate mixing. You could also use this same feature to remove bleed from recordings where multiple people are on separate mics in the same room. De-bleed and separating speakers to their own tracks would each be extremely valuable new features for you to add.

I would love the ability to choose how many seconds I want long deadair shortened to, or least a few options. I prefer to limit silences to 2 seconds instead of 5.