rating

Triggered when an AI-powered quality rating is generated for a call. Ratings evaluate call quality across multiple criteria and provide actionable feedback.

When This Event Fires

• The AI has finished evaluating a call's quality
• The rating results are stored and ready for retrieval

Payload

Field	Type	Description
event	string	Always "rating"
call_id	integer	Internal call ID
overall_score	integer	Overall quality score
criteria_scores	array	Scores for individual evaluation criteria (see below)
highlights	array of strings	Positive aspects of the call
improvements	array of strings	Suggested areas for improvement

Criteria Score Fields

Each item in criteria_scores:

Field	Type	Description
name	string	Criterion name (e.g. "Politeness", "Problem Resolution")
score	integer	Score for this criterion
feedback	string	Specific feedback for this criterion

Example Payload

{
  "event": "rating",
  "call_id": 12345,
  "overall_score": 8,
  "criteria_scores": [
    {
      "name": "Politeness",
      "score": 9,
      "feedback": "The agent maintained a friendly and professional tone throughout the call."
    },
    {
      "name": "Problem Resolution",
      "score": 7,
      "feedback": "The issue was addressed but a follow-up action was left unconfirmed."
    },
    {
      "name": "Product Knowledge",
      "score": 8,
      "feedback": "The agent demonstrated strong knowledge of the product features."
    }
  ],
  "highlights": [
    "Excellent greeting and rapport building",
    "Clear explanation of pricing options",
    "Proactive suggestion of relevant add-on services"
  ],
  "improvements": [
    "Confirm next steps before ending the call",
    "Ask if the customer has any additional questions"
  ]
}

Related Events

• speech — the transcription the rating is based on
• summary — AI summary for the same call
• incoming_call_end / outgoing_call_end — the call that was rated