Skip to content

AI insights

An insight is a specific piece of information the AI extracts from a conversation — the reason for the call, the product discussed, a customer-satisfaction score, a missed sales opportunity. Your organization defines which insights to extract; MiaRec fills them in on every analyzed conversation.

View insights on a conversation

  1. Open a conversation from the Conversations list.
  2. Click the Analytics tab.

Insights are grouped into collapsible sections. The groups and fields you see depend entirely on your organization's configuration; a typical setup includes groups such as:

  • CX metrics — customer-experience scores like CSAT or NES, each with the AI's reasoning.
  • Call reason & outcome — why the customer called and whether the issue was resolved.
  • Sales metrics — for example, missed sales opportunities and sales objections, with quotes, timestamps, and coaching tips.
  • Custom insights — any other data points your organization extracts.

Most insight values come with the AI's explanation of how it reached the answer, often with supporting quotes from the transcript, so you can verify the result.

The Analytics tab groups the insight values for one conversation — for example CX metrics (CSAT, NES), call reason and outcome, and sales metrics with objection quotes and coaching tips — with the Topics section closing the tab.

Edit an empty insight value

Insight values are stored in custom fields, and fields the AI left empty show [Click to edit]. If your role permits it, click the value to enter it manually — useful for fields that are sometimes filled by hand rather than by the AI.

Use insights beyond one conversation

  • Search — each insight field appears as a Call - <field name> parameter in advanced search, so you can find, say, every conversation with a low CSAT.
  • Dashboards — numeric insight fields enabled for dashboards get their own dashboard tabs (for example a CSAT tab). See CX metric dashboards.
  • Reports — insight fields are available as report columns, including per-field explanation and average/total columns. See the column reference.

Note

Defining new insights — the prompt, the response type, and the destination field — is an administrator task. See Creating custom insights in the Administration Guide.

  • Topics — the topic detection results shown at the bottom of the same tab.
  • Summaries and Sentiment — the other AI results on the detail page.