# I was thinking maybe embed voiceprints into neuros memory, so she can remember speakers. Another idea is that using voiceprints she could detect the speaker, this way maybe she would have more conversation context in group calls.