Capture Stage
VAD-triggered chunks sent to Whisper
Raw chunk data will appear here...
VAD Settings
These settings apply when recording starts. They cannot be changed mid-recording.
Processing Stage
Segmentation, clustering, and speaker attribution
Edit Job
Delete this job?
No participants yet
Transcript will appear here when you start recording...
Processing in progress. Settings are locked.
Model settings are locked after processing. Clone this job to change them.
Models
Embedding:
Segmentation:
Segmentation Params
Speaker Matching
Configured per-input in topbar
Similarity:
Confidence:
Max Speakers:
Conversation Boosting
Boost Factor:
Eligibility Rank:
Max Margin:
Skip if Confident:
Min Similarity:
Min After Boost: