No servers.
Transcription, speaker identification and language detection run on the phone. There is no account, and nothing to upload.
Name a voice once.
Name a speaker in one interview and the app recognises that voice in the next. The voiceprint stays on this phone with everything else.
Share the file, not the access.
Fix anything it got wrong, then send the transcript as text, Markdown or Word. What you share is the only way it leaves.
Interviews · Meetings · Any conversation worth keeping