Introduction

The Auto QA component in wolkvox Studio allows you to automate the quality analysis of voice and chat interactions using artificial intelligence, based on a survey (matrix) created in Quality Analyzer.

Its purpose is to select a percentage of interactions, evaluate them using the questions defined in the matrix, and save the results into variables that you can use later (for example, to send email notifications with the “wolkvox notification” component or for integrations via CRM + Webhook).

This component is only available for the Routing Point type: CRM + Webhook + Cron + VOC Analytics.

Requirements

Before configuring the component in Studio, make sure the service is enabled in the operation:

1. wolkvox Manager Configuration.
2. “VOC” tab.
3. In the VOC configuration section:
   • Enable “Activate text analysis.”
   • Enable “Activate Speech Analytics.”
   • In the “VOC - wolkvox studio - Business logic” field, select the Routing Point where you will execute the Auto QA component.
4. Click “Save.”

Configuration

1. Open the component's configuration panel by double-left-clicking on it.
2. The following fields are displayed:
   • Language: Select the language of the conversations to be evaluated.
   • % of interactions to evaluate: Define the percentage of interactions that will be scored (e.g., 10, 25, 100).
   • Select the survey to score for this interaction: Select the survey (quality matrix) you created in Quality Analyzer. This survey defines the questions/criteria used to evaluate each interaction.
   • Critical error unit: Variable where the result associated with the critical error “unit” will be saved.
   • Critical error opportunity: Variable where the result associated with the critical error “opportunity” will be saved.
   • Average score (variable): Variable where the average score result of the interaction will be saved.
   • HTML table result: Variable where the result will be saved in HTML format, useful for:
     • Sending via email (for example, with wolkvox notification).
     • Rendering on a front-end or integrating into an external system.
   • Use the latest artificial intelligence language model (More expensive): Enable this if you wish to run the analysis with a more advanced model (at a higher cost).
   • Finally, click “Save” to apply the changes.