OVERVIEW
Wonda provides a powerful tool for analyzing discussions with specific characters, enabling in-depth assessment of participants’ mastery against specific criteria you define.
This feature is particularly valuable for providing feedback to participants at scale and assessing simulation performance across participants.
Based on the inputted criteria, this feature provides:
• A summary of the trainee’s overall performance and improvement areas
• Trainee’s specific strengths and feedback for each criterion
• A cumulative report with performance assessments for the last 10 sessions
Image 1: A report generated based on a participant’s discussion with a negotiation simulation
The feature works with existing discussion logs, and thus can only be used with experiences that contain discussion logs.
Limitations of the beta version that will be addressed in the future:
The experience cannot include more than one conversational avatar.
Participants will not be able to view their personalized feedback report unless downloaded and shared by the trainer.
The assessment report will generate up to 10 reports based on the last 10 discussions with the character in an experience.
TABLE OF CONTENT
In this article, you will discover:
How to access the AI Assessment Wizard
The AI Assessment Wizard is currently in beta mode, meaning it only works with experiences involving a single character.
You can find it here:
[Experience > Analytics page > ✨ AI Assessment (beta) tab]
Gif 1: Accessing the AI assessment feature
Defining Assessment Criteria
To generate reports, you will need to provide criteria that will serve as the basis for the generative model to create a personalized assessment.
You can provide these criteria via the AI Assessment Wizard, accessible by clicking on [Create a New Assessment], and then on [Edit].
You will need to provide:
A title and description for your assessment
Between 3 to 6 criteria, each with a title and description
Image 2: Filling out the criteria
Based on the criteria and descriptions you provide, the model will evaluate the individual's performance for each discussion.
Note: The assessment is performed by the
gpt-4o-2024-05-13model.Once the criteria have been entered, you can save them by clicking the [SAVE ASSESSMENT] button.
Image 3: Save your criteria
Understanding the Assessment Reports
To generate reports, simply click the [GENERATE THE LATEST REPORTS] button.
This will generate up to 10 (beta) reports based on the last 10 discussions with the AI Character.
Image 4: Reports generated by the AI Assessments
Once the reports have been generated, you will be able to preview them directly in the ✨ AI Assessment (beta) tab.
Simply click on a card to open the detailed report.
Image 1: A report generated based on Terrance’s discussion in the negotiation simulation
You now have all the basics to use the AI Assessment feature.
As it is new and currently in beta, we would greatly appreciate your feedback on what you like and how we might improve it. To do so, simply use the feedback option provided in the interface.
You may also be interested in exploring Analytics Dashboards for more in-depth insights.