local coaching nudges
a 3-minute cooldown applies per coaching nudge type, so you won’t get spammed. each prompt is short and dismissible.
which speaker is coached
- macos - nudges are computed only for an explicitly marked speaker or the sole mic speaker in a normal dual-source online call. when several people share the mic and nobody is marked as you, miniti does not attribute coaching to the room.
- ios - capture is mic-only, but live transcription can still separate people sharing the microphone. nobody is assumed to be you, so speaker-specific coaching waits until you mark your speaker.
when they fire
on macos, nudges appear in the floating recording surface while miniti is frontmost. when another app is active, or the floating surface is off, miniti uses a system notification. on ios, they remain background notifications.enable them
- macos - settings (
⌘,) → notifications → meeting nudges → toggle nudge me when i’m monologuing and nudge me when i’m using too many fillers - ios - settings → notifications → meeting nudges → same two toggles
privacy
- monologue, filler, and Sales-conversation detection run locally from the live transcript without an extra model request
- incisive questions come from the enabled Questions insight view and follow the normal ai insights data flow
- system delivery uses a local
UNNotificationRequestscheduled by miniti
tuning
the thresholds (180 words, >8 fillers/min, 60-second window, 3-minute cooldown) are fixed in this release. if you want different values, use the support form with your use case.related
- speech coaching - in-app metrics during the meeting
- filler words - how filler detection works per language
- delivery metrics - how to think about monologue length

