Skip to content

AI QA

For how AI QA fits together with QA forms, manual reviews and the QA statistics, see How Quality Works.

Function description

Daktela AI QA is an advanced tool that leverages artificial intelligence to automate the quality assurance (QA Reviews) process in contact centres. By analysing communications across various channels, AI QA ensures consistent service quality and saves time for your team leaders. It integrates with your existing QA forms, allowing reviews to be evaluated by AI while simultaneously being conducted manually by a team leader. The team leader can later adjust AI-generated QA reviews if any corrections are needed. AI QA Reviews generated by an AI Agent are available in the Listings module under QA Reviews, just like standard ones reviewed by your team leader.

Info

AI QA can only process text. To use it with calls, the calls must be recorded (When to record calls automatically in the queue settings β€” recording is off by default) and Smart Call Transcript (Speech-to-Text) must be enabled on the queue.

Licensing and Setup

Licensing

AI QA consumes 15 Power Pack tokens per question per activity analysed β€” for example, a 10-question form uses 150 tokens for each activity. Administrators can manage Power Pack licence packages in the Licensing Module under the Add-ons Tab. Consumption is shown in the AI Hub.

Adding a Power Pack package on the Add-ons tab of the Licensing module

Setup and Configuration

  • Click Change in the AI functions column of the queue in Settings β†’ Routings β†’ Queues β€” see Queue Basics β†’ AI Functions.
  • Enable the Use AI QA Reviews checkbox.

  • Choose the AI QA Form that you want to use. If you do not have an existing QA Form, you can create one in Settings β†’ Workflows β†’ QA Forms.

    • When creating a new QA Form, do not forget to check AI QA Form.

    • In the Form Builder, edit each question and provide a detailed description for the AI. Choose a question Type, fill in the Title and Description β€” those are used in the instructions given to the AI. If the type has Options, you can define them below. The option descriptions are also used in the instructions given to the AI.

      Editing a QA question with title, description and answer options for AI QA

    • After you have finished creating new QA questions, you can drag and drop them into the form.

      Dragging QA questions into the form in the Form Builder

    • You can then set Conditional Evaluation for each question in the form. If you need a question to be evaluated based on the result of another question, you can set this by clicking on the gear icon.

      Setting conditional evaluation for a question with the gear icon

  • Specify the AI Language for the output.

  • If you do not want to analyse every interaction, define AI QA conditions and AI QA Frequency (from All activities down to every 30th activity). Reviewing a sample keeps the token cost under control.

Warning

The Linear Scale type of question is not available for AI QA.

On-demand AI QA

You can manually trigger the generation of AI QA for a specific activity. This can be done in the activity details using the On-demand AI button. If it's a call without a generated Smart Call Transcript, the system will generate both the transcript and the AI QA. Before confirming, you will see how many tokens (and potentially how many minutes for the transcript) will be consumed. The On-demand AI button is enabled only if AI QA is activated for the queue associated with the activity. The user also needs the AI ondemand actions permission in their user access. This feature can be useful in two scenarios:

  1. The activity is from an earlier date when AI QA was not enabled.
  2. You have updated your AI QA setup, but an older activity still uses the previous configuration.

Generating AI QA on demand from the activity detail