v0.6.0-beta15
Last build started showing you the voices it was hearing. This one lets you answer it.
Tell Minutiva who it’s hearing — mid-meeting
The roster above the live transcript is now clickable. Where it guesses a name, you can confirm it (Yes, this is Marisa) or wave it off (Not Marisa). Confirmed names lose the question mark and stay put for the rest of the meeting.
Voices it doesn’t recognise get Name this voice… instead. There’s no suggestion to accept for someone who isn’t in your People list yet, so you just type who they are — they’re added to People when you stop.
Your answer sticks, even when the app changes its mind
Minutiva re-works out who’s who every thirty seconds as more of the meeting arrives, and the voices can get renumbered in the process. Your answers are pinned to the actual speech you were looking at, not to a position in a list, so they follow the right person through every one of those passes — and they’re still there in the finished transcript after you stop.
If the app genuinely can’t tell where a voice went — it split a speaker in two, say — it won’t quietly guess. The name drops back to a suggestion you can accept in the speaker panel afterwards. A missing label is easy to fix; a confident wrong one is the kind you don’t catch.
Also
- Answered a chip by mistake? Undo is in the same menu.
- Saying “not Marisa” only rules out Marisa — if the app later thinks that voice is someone else, it’ll still say so.
Auto-update
Existing beta channel testers get this within 24 hours. Or trigger now: Minutiva Private → Check for Updates…