Call Analysis

Overview

Sulus's call analysis runs automatically after every call ends and produces several types of results that are included in the end-of-call report. Here's a comprehensive breakdown of how it all works together.

As soon as a call ends, analysis is triggered in the background and typically completes within a few seconds. Results are attached to the call record and can be viewed in the call instance dashboard or retrieved via the API.

The analysisPlan in your assistant configuration controls three main outputs:

Summary — A concise summary of the call, stored in call.analysis.summary. The default prompt summarizes the call in 2–3 sentences. You can customize or disable it:

{
  "summaryPrompt": "Custom summary prompt text"
}

Structured Data — Specific data extracted from the call transcript, stored in call.analysis.structuredData. You define the shape of the data using a JSON Schema:

{
  "structuredDataSchema": {
    "type": "object",
    "properties": {
      "field1": { "type": "string" },
      "field2": { "type": "number" }
    },
    "required": ["field1", "field2"]
  }
}

Success Evaluation — A determination of whether the call was successful, stored in call.analysis.successEvaluation. You can choose from several rubric types:

The end-of-call report is delivered via a server webhook event of type end-of-call-report. It includes the call analysis alongside other artifacts:

{
  "message": {
    "type": "end-of-call-report",
    "endedReason": "hangup",
    "call": { },
    "artifact": {
      "recording": { },
      "transcript": "AI: How can I help? User: ...",
      "messages": [
        { "role": "assistant", "message": "How can I help?" },
        { "role": "user", "message": "..." }
      ]
    }
  }
}

The analysis object (containing summary, structuredData, and successEvaluation) is also included in the report payload.

You can combine a custom prompt with a rubric for more detailed evaluation instructions:

{
  "successEvaluationPrompt": "Evaluate the call based on these criteria:...",
  "successEvaluationRubric": "Checklist"
}

To disable the summary or success evaluation, set the respective prompt to an empty string:

{
  "summaryPrompt": "",
  "successEvaluationPrompt": ""
}

In summary, Sulus's call analysis automatically generates a summary, extracts structured data, and evaluates call success after every call. All results are accessible via the call record in the dashboard or through the end-of-call-report webhook, and each component is fully customizable through your assistant's analysisPlan.