After the call: summary, tasks and export
If saving to History was on, the call becomes a document: both sides of the conversation, a summary of what was agreed, and the tasks that follow from it. Audio is never part of it.
Where it is kept
In a database on your machine. Nothing is uploaded unless you turn cloud sync on, and the search across past calls runs locally, in both languages at once.
The summary is produced once, when the session ends, by sending the text of the transcript for summarisation. If you would rather that did not happen, Settings → Privacy has the switch for it, and the transcript is still saved.
Naming the voices
The recogniser separates voices but does not know who they are. Click a speaker label and type a name — it is applied to every line that voice said, to the summary and to any export you make afterwards.
Exporting
Markdown or plain text, with a checkbox for whether to include the translation. Without it you get the conversation as it was spoken, which is usually what you send to the people who were on the call; with it, what you send to the people who were not.
The export carries the summary, the task list and the timestamps, so it can be pasted into a tracker without editing.
Read next
- TroubleshootingThe problems people actually hit: no system audio, text that stops arriving, the capture-restarted notice, minutes running out, and where the log file is.
- Meeting contextSet up a preset for a meeting: your role, the goal, the counterpart, the tone, a glossary of terms, and documents the answers should quote.