Skip to main content

What are Roleplay Session scene types?

Learn about the three Roleplay Session scene types in Synthesia — scripted, roleplay, and feedback — and when to use each one.

Written by Meg Farley

📚 Need the big picture first?
​
​
See What are Roleplay Sessions? for how the whole experience fits together.

Roleplay Sessions are built from scene types that work together to create a complete learning experience. Scripted scenes teach or model the skill, and the AI Roleplay activity lets Learners practice with an Avatar and see automated results afterward.


The two scene types

A Roleplay Session combines two scene types, each doing a different job.

Scripted scenes

Scripted scenes teach the skill and show what good looks like. An instructor or AI avatar demonstrates the technique using concrete examples, real customer scenarios, or best practices.

Scripted scenes are passive — learners watch and absorb, they do not interact. Use scripted scenes to set up context before the roleplay so learners know what they are practicing.

Scripted scenes

AI Roleplay scene

The AI Roleplay activity is where Learners practice and get scored. It's a single scene with two views — Roleplay and Results — that always move, get selected, and get deleted together.

Roleplay view

This is where the practice happens. The Learner has a real-time conversation with an Avatar who plays a specific role and responds authentically to what the Learner says. The Avatar follows the scenario you define, such as a customer calling with a complaint, a prospect with objections, or a teammate asking for feedback.

Roleplay scene

Results view

This appears after the Roleplay ends. The coach walks the Learner through what they did well, where they can improve, and how they performed against the skill criteria. Learners see their overall result, their score for each skill criterion, and detailed coaching comments tied to the conversation.

Feedback scene
​


How the scene types work together

The typical Roleplay Session flow is: Scripted → Roleplay → Results.

Learners watch the teaching content, practice with the Avatar, then get scored feedback. This flow helps Learners see a model answer, try it themselves, and get specific coaching.

✍️ A standalone AI Roleplay activity also works. This suits assessments, where you do not want to give away the model answer before the Learner attempts the conversation.

Scene limits

  • One AI Roleplay activity per video — After you add one, you cannot add another. Its Roleplay and Results views always move, get selected, and get deleted together, and no other scene can be inserted between them.

  • Multiple scripted scenes — You can add multiple scripted scenes before the Roleplay, but keep the scripted content short so Learners get to the practice quickly.



💬FAQs

Do I need both scene types?

No. The typical sequence is scripted scenes followed by the AI Roleplay activity, but a standalone AI Roleplay activity also works — especially for assessments where you do not want to give away the model answer.

What order should the scenes go in?

Scripted scenes come before the AI Roleplay activity, so Learners are informed and shown the skill before they are asked to practice it. Its Results view appears automatically after the Roleplay ends.

What is the difference between a scripted scene and a AI Roleplay activity?

A scripted scene is an Avatar narrating a script you write. The AI Roleplay activity is a live, two-way conversation where the Learner speaks and the Avatar responds, followed automatically by the Results view.

How many AI Roleplay activities can one video have?

One. The option to add another is disabled once one already exists in the video, since its Roleplay and Results views are always part of a single activity.

Can users customize how a roleplay or feedback scene looks?

Only through the options exposed in the design panel, such as colors or background. Free-form canvas editing is not available for these views.

Can any avatar be used as an interactive avatar?

Most can. All Express-3 stock Avatars and almost all Express-2 stock Avatars can be used as interactive Avatars. Pick a background from the Avatar picker, or select Customize Avatars to set a specific outfit and space from a prompt.

Can I use the same Avatar in my scripted scenes and the Roleplay?

Yes. Because the interactive pool now matches the full stock range, the same Avatar can narrate your scripted scenes and then coach live in the Roleplay — useful for a Coach character who introduces the Session and then coaches the conversation.

Did this answer your question?