Skip to content

Conversations and Data Sources

In MiaRec, a conversation is a single customer interaction that can be analyzed to produce insights.

Depending on your deployment, a conversation may be:

  • a voice call
  • a chat conversation
  • an email thread
  • a ticket with message history

In some parts of the product UI, you may still see “call” used for voice-specific views. The broader term “conversation” is used here to support omni‑channel analytics.

What makes a conversation analyzable

To analyze a conversation, MiaRec needs:

  1. Content
  2. voice calls: a recording that can be transcribed
  3. text channels: message/thread text
  4. Metadata (recommended)
  5. time and duration
  6. direction (inbound/outbound), where applicable
  7. agent/team/queue identifiers
  8. tags and other business context

Metadata enables segmentation and filtering in dashboards and search.

Common data sources (examples)

MiaRec deployments often integrate with:

  • contact center platforms (voice recordings and call metadata)
  • chat platforms (messages, timestamps, participants)
  • ticketing systems (ticket threads, status changes)
  • email systems (thread text and headers)

The exact connectors and ingestion methods depend on your environment and are described in the Platform Setup & Operations Guide (for operators) and the Administration Guide (for tenant admins).

Why “conversation” is the right unit

Using a common unit across channels provides consistency:

  • the same insight (e.g., topic, sentiment) can apply to calls, chats, and tickets
  • dashboards can show trends across channels
  • search filters work the same way (e.g., “High churn risk”)

Next: transcripts and threads

To understand how MiaRec turns content into analyzable text: