In NeuroSwitcher 0.8.3 recording a meeting takes almost no effort: starting, transcribing, naming the people and writing the summary now happen on their own.
- Names instead of “Speaker B”. The people on the other side are told apart by their voices — on your own computer — and each person stays one voice for the whole meeting.
- Checked by ear. Five clips of each voice from all through the meeting (more on request), ▶ on every line. A wrong clip is moved to a person by name.
- A voice book. Confirmed voices are remembered: in the next meeting with the same people their names go in by themselves; when unsure — “maybe Andrey?”.
- Neuron offers to record. A call starts in Slack, Zoom, Teams, Telegram or a browser — Neuron asks: “Record the meeting?”
- Transcript right after Stop. The recording is named after the call app and the time; the transcript and the summary are ready by the time you open them. The summary lists who took part.
- Slack. Slack's audio makes voices sound alike — for such calls the speakers are told apart by turns and pitch.
Voice prints are computed on your PC and never leave it. The voice model (27 MB) is downloaded once.
Also in 0.8.3: links in answers about meetings open the transcript at the right minute, ✨ adds the subject to a recording's name, and “Transcribe again” is right in the transcript window.
More: https://neuroswitcher.app/meeting-recorder.html#voices