Summary is concise, AI-generated overview that captures the main events, actions, and outcomes of a customer-agent interaction. It highlights the sequence of events, key customer issues, actions taken by the agent, and any next steps or resolutions agreed upon in a case. A summary helps contact centre teams quickly understand what happened in each case without reading the entire transcript, enabling faster review and efficient follow-up.

​

Topics is an advanced, AI-powered feature that automatically scans entire customer-agent interactions across both text and voice channels to detect and classify every key issue, experience, or action discussed in a case beyond the primary contact drivers. This capability recognizes multiple relevant topics, including customer problems, expressions of dissatisfaction, and specific agent behaviors (such as transfers, holds, or mutes).

By analysing every message within a conversation, Summary and Topics provides a detailed, structured view of all critical themes within a case, ensuring a more complete understanding of what transpired during each case.

​

In the Case Analysis Widget, you can view both Summary and Topics. The Topics section displays AI-detected topics discussed during the conversation, while the Summary provides a detailed overview of the key events covered.

Expanded view of this section shows all topics, ordered by different Topic Types along with the messages they are detected on:

Different Topic Types

Topics is a feature in Conversational Analytics that identifies multiple key topics discussed within a single customer conversation. Unlike models that capture only the primary contact driver, this feature analyzes the entire interaction to extract deeper context through five specific topic types. The following are the multiple types of topics that are displayed:

1. Customer Issue Detail clearly captures the specific, customer issue or query related to the main contact reason.​

2. Customer Dissatisfaction captures strong negative emotions or concerns expressed by the customer, such as frustration, anger, legal threats, or suspicion of fraud.​

3. Agent Transfer Reason tags captures the specific reason an agent explicitly transferred a call.​

4. Agent Mute Reason captures the specific reason an agent explicitly put a call on mute.​

5. Agent Hold Reason captures the specific reason an agent placed a customer on hold, identified by an explicit hold notification from the agent.​