Call Score API
The Call Score API assesses conversation quality and participant performance at scale.
The Call Score API enables you to assess conversation quality and participant performance at scale. It provides a numerical score for each conversation, making it easier for business leaders to identify and compare relevant conversations. You also get a detailed breakdown of the score based on criteria suited for the type of conversation. Because the API is powered by generative AI, it offers clear and concise explanations for each criterion’s score, enabling participants to understand their strengths and areas for improvement.
You can access Call Score in the following ways:
- API Integration: Use the Call Score API to retrieve call scores with your application. You can build your application interface to display the scores using the API’s responses. To use the API, see Get started.
- API with Insights UI: Use the Call Score API to generate call scores and leverage the pre-built Insights UI. This UI provides an out of the box interface to display scores along with other conversation insights such as summary, sentiment analysis, next steps, and Q&A. For more information, see the Insights UI .
- CRM integration: Push call scores directly to your CRM using the call score API and use CRM analytics tools to generate reports based on call scores.
Use Cases
- Sales Performance Assessment: Chief Revenue Officers and RevOps Managers can objectively evaluate their sales team's performance using Call Score. This eliminates the need to listen to recordings of sales calls.
- Contact Center Quality Evaluation: Contact Center teams can assess the quality of all calls using Call Score. Supervisors and QA managers can automatically score each call without having to listen to recordings.
- Positive Communication Culture: Enterprises can ensure employees maintain a polite, inclusive, and empathetic communication style during meetings with Call Score, promoting a positive communication culture.
Get Started
To get started, use the following steps:
- Authenticate and generate an access token
- Process a conversation using
- Async API
- Streaming API
- Get call score via Conversation API
Here are the steps in detail:
-
Process a conversation
-
Use the Async API to process a recorded text, audio, or video conversation. You can use the following attributes to use Call Score for your business use case. To learn more about getting started with other Async APIs, please refer here.
Here is the sample Async Text API request
POST /v1/process/text { "name": "Postman Default Meeting (Sales/Demo)", "features": { "featureList": [ "callScore", //Adding callScore to the featuresList kicks-off the call score process. "insights" ] }, "conversationType": "sales", "metadata": { "salesStage": "qualification", "prospectName": "Audio File LLC" }, "messages": [ { "payload": { "content": "So that was great to just kind of went through and be prepared for what you were looking for." }, "from": { "userId": "[email protected]", "name": "Agent John" } } ] }
-
Use the Streaming API to process a real-time conversation. You can use the
generateCallScore
action in the Streaming APIstart_request
which triggers call score processing once the streaming session ends. To learn more about getting started with Streaming API, refer here.Here is the sample Streaming API start request
{ type: "start_request", actions: [ { name: 'generateCallScore', parameters: { conversationType: "general", salesStage: "general", scorecardId: "5733078868164608", prospectName: "Realtime Call Score", callScoreWebhookUrl: "https://webhook.site/dc3907b1-429a-466a-97e2-2eae00b752da" } } ] }
-
- Use the Conversation API to GET Call Score API endpoint.
Here is a sample Get Call Score request:
GET /v1/conversations/{conversationId}/callscore
Call Score Concepts
Criteria
Criteria are detailed measures used to evaluate specific qualities of the speaker in a conversation. They are essential components of a scorecard, helping systematically assess communication quality and performance of the speaker in a conversation. Each criterion is uniquely identified and can include managed criteria provided by the system or custom criteria created by the user. These criteria measure specific qualities or performance metrics, such as empathy, clarity, and problem resolution. They consist of a name, a checklist of evaluative questions, assigned priorities to highlight critical areas, and optional descriptions and tags for context and organization. To learn more about criteria, refer here.
Scorecard
A scorecard in the Call Score API serves as an abstraction layer combining various criteria for evaluating conversations. It can include custom criteria, managed criteria provided by Symbl.ai, or a combination of both. This flexible approach allows you to tailor the evaluation framework to meet specific business needs, ensuring a comprehensive and relevant assessment of call quality. To learn more about scorecard, refer here.
Updated 5 months ago