Conversation Intelligence
Reads call transcripts and email threads to extract summaries, sentiment, action items, risk flags, and competitor mentions.
Conversation Intelligence
What it does. Processes call transcripts and email threads with AI language understanding so nothing important gets lost in a recording no one re-reads.
Signals it reads. Raw transcripts and email subject/body (treated as untrusted input and sandboxed against prompt injection).
What it produces. A summary, sentiment, topics, action items, risk flags, and competitor mentions written back onto the record; a Timeline entry; Actions for the action items; and Risk tasks for high-severity flags.
Autonomous vs on-request. Autonomous — it sweeps for unprocessed transcripts and emails every few minutes.
Your control. Enable/disable in Admin Console → AI Copilot & Agents → Agents. Extracted action items surface as Actions you can accept or close.