Documentation

AI call evaluation scenarios

Author: Łukasz Strumiński

Created: Jun 20. Updated: Aug 1.

In the Thulium system, you can fully automate the assessment of the quality of customer conversations by creating personalized assessment scenarios. These scenarios consist of a set of evaluation criteria that form the basis for subsequent analysis of the conversations conducted by the Agents in question. This approach makes it possible to dispense with manual analysis or limit it only to selected cases, significantly increasing the scale and efficiency of the quality control process. To prepare such a scenario, select AI Assistant → AI call evaluation scenarios from the Administration tab.

Next, select one of the ready-made assessment scenarios, or create your own by clicking on .

After selecting a scenario, a list of criteria that make up the scenario will appear. Each criterion is formulated in a zero-one manner (fulfilled or not), which allows an objective and reproducible evaluation of the team's work.

If the scenario suits our needs then save it by clicking . On the other hand, if we need to add some criteria then we use the button. Then, from the list of predefined criteria, we select the one we are interested in by approving the selection with the button:

The given criteria in the list in the scenario can be arranged according to the order you expect by holding down . Removal of a criterion is possible by selecting .

The created evaluation scenario will appear in the list of available scenarios in the left section of the module:

If the need arises to create a new scenario, but based on an existing one, then after selecting such a scenario, click and proceed as before when adding a new one.

A scenario that is not used in the rules can be deleted by clicking .

Automate quality control and focus on service that delights! Create call evaluation scenarios and define what really matters in customer interaction.