Quick answer: Add the lesson from Curriculum > Add Lesson Material > Exam > AI Interaction Review, then pick a rubric so Smart AI grades the recordings. Learners upload or record their conversation, and you review the results under Product > AI Interaction Review.

AI Interaction Review grades a conversation between two people. Typical cases are a sales person talking to a buyer, or a customer service agent talking to a customer.

The learner submits the conversation as audio, either by uploading a file or by recording directly. Smart AI then transcribes it, scores it against a weighted rubric, and writes a summary report with strengths, areas for improvement, and recommendations.

You keep the final say. Every AI score can be edited before you publish it.

Open the course you want, go to the Curriculum tab, find the session you want the lesson in, and click + Add Lesson Material.

The picker groups materials into Media, Exam, and AI Persona. Choose AI Interaction Review, which sits in the Exam group next to Assignment, Quiz, Self-Assessment, and Survey.

Fill in the Lesson Title, then use the Description editor to explain the scenario you want learners to record.

Be specific here. The description is where learners find out who they are supposed to be talking to and what situation they are roleplaying, which directly affects how well their recording scores.

This is the step that decides how the lesson is graded.

Under Select Rubric, click the field to open the rubric picker. It states the rule directly: By selecting a rubric, grading will be assisted by AI.

In the picker you can:

Choose a rubric, then confirm with Select Rubric.

If you leave this empty, the lesson stays on [Manual Grading], described in the form as "Educator will manually grade and provide feedback". The AI will not score anything.

Under AI Interaction Review Setting, tick Set Passing Grade if learners need to reach a minimum score.

Under Material Setting you can tick Set this as a required lesson and enter a Duration in hours, minutes, and seconds.

Click Save when you are finished. Draft keeps the lesson unpublished while you keep working on it, and Discard Changes throws away what you have entered.

In your Eduqat Dashboard, go to Product, then AI Interaction Review in the panel that opens.

The page lists every course that has reviews, with four columns: Course Name, AI Interaction Review (how many lesson materials are being reviewed), Date Created, and Educator. You can search by course name, sort by course name, and change how many rows show per page.

If you see "No AI interaction reviews yet", check the other product type tabs before assuming there is no data. The tabs are Scheduled Session, Self-Paced, Cohort, and Digital Download, and each one holds its own list.

Click Details at the end of the course row.

The AI Interaction Review Details page shows the course, a Search lesson name box, and the lessons being reviewed. Each row shows Lesson Name, Submission (how many submissions it has), Creator, and Date Created.

Click View Detail on the lesson row.

This is the main review page for that lesson. At the top you can filter by date range and by group, and you get five figures at a glance:

Below the figures is Learner Progress Evolution, which tracks performance across multiple graded submissions.

Two dropdowns control it. The first switches the view, for example Per Attempt. The second picks whose data you see, for example All Learners.

The chart plots one line per rubric criterion, so you can see which specific skill improved and which did not. The criteria come from the rubric you chose in Step 3, which is why the legend changes from lesson to lesson.

Under the chart, the Overall Summary box repeats the headline average for the group you selected.

The table at the bottom lists everyone assessed, with Name, Group Name, Uploaded Count, Avg Score, First Submission, and Last Submission.

Click the three dot button at the end of a learner's row. It offers three actions:

View Submission opens the learner's files, grouped by date and newest first, with a date range filter and a sort control.

Each file shows its name, the time it was submitted, and its size, with an audio player and a download button. Expand Audio Transcription to read the conversation as text instead of listening to it, which is usually faster when you are checking many submissions.

On the right you get the learner card, a Select Attempt dropdown that shows which attempt you are looking at and whether it is Graded, a View Analytic button, and the Final Score with its status.

Use the arrows next to Learner's Name to move to the next learner without going back to the table.

View Analytic opens the Summary Report for that learner.

The report has a headline paragraph followed by three sections: Strengths, Areas for Improvement, and Recommendations. At the end it is signed Generated by Smart AI with the date and time it was produced.

Three controls sit at the top of the report:

Back on the submission page, click Edit Grading. The header buttons change to Save as Draft and Publish, which is how you know you are in edit mode.

The grading panel has two parts:

Click Show Score Definitions on any criterion to see what each score level means before you change a number.

When you are done, use Publish to publish the result, or Save as Draft to keep your changes without publishing. A published score carries a Published badge next to the Final Score.

If a submission does not appear or a score will not publish, reach out through the live chat button in the bottom right corner of your screen and we will look into it with you.