Keeping an eye on the repo for speaker diarization/attribution.