Convaa Gather is live.
Most meeting tools treat the call and the write-up as two separate places. You hold the meeting in one app, then go find the transcript in another, then paste the useful parts into wherever your team actually works. The handoff is where the context goes missing.
Gather removes the handoff. You start a call from the thread you are already in. When the call ends, the transcript is in that same thread, next to the decisions it belongs to.
Cameras work. Screen share is not granted at any setting, and video is never recorded. The only thing captured is audio, and only because that is what the transcript is made of. That is a deliberate boundary rather than a limitation we have not got round to: a camera in a recorded room is a different conversation from a microphone in one, and sharing a screen into a room whose notice says it records voice is a different thing again.
The part that decided the architecture is self hosting. Some teams cannot let meeting audio leave their network. Legal, health, public sector, defence suppliers, anyone under a data residency rule written before anyone had heard of a meeting bot. For those teams, the whole capture path can run on their own hardware: the media server, the transcription, the storage. The audio never has to cross the boundary.
If your meeting notes keep ending up somewhere nobody reads, this is the shape of the fix.