Who is speaking

ozen recognizes people by their voice. Someone keeps the same label for the whole meeting, and once you’ve told ozen their name, it recognizes them in future meetings too.

New voices appear as S1, S2, … until you name them.

Naming someone

1. Click the name on any line they said, like S2.
2. Pick their name from the list, or type a new one and press Return.

Within a few seconds ozen updates its memory of that voice and relabels every line in the meeting, including earlier ones. A couple of names per person early in a meeting usually fixes the rest.

Made a mistake? Click the name again and choose Clear tag.

Review

When ozen isn’t sure who said something, it puts an orange ? next to the name. The Review button counts those lines. Press it and ozen jumps to the one it’s least sure about and asks who said it.

Review only asks about the last 10 minutes, while you still remember.

The fastest way to better names

Answering Review teaches ozen much faster than naming random lines, because it asks exactly where it’s confused. In a test with four similar voices, 12 answers to Review got every line right.

Ignoring a voice

A TV, a video next to your Mac, or a podcast isn’t part of your meeting. Click its name and choose:

• Ignore this voice for that voice, or
• Ignore all lines by S3 to clear everything it said nearby at once.

Those lines turn grey, and ozen stops writing that voice down. To undo, name the line as a person, or choose Clear tag.

People talking at once

When two people talk over each other, ozen separates them, and each gets their own line. On the Timeline, their bars overlap.

Echo from your speakers

If your call plays through speakers, the microphone hears it too. ozen notices that the “room” voice is actually the call playing back and drops the duplicate. The same goes for your computer reading text aloud.

Someone in the room talking over the call still gets their own line, because their voice doesn’t match what was playing.