Roleplay Suggestions: Build with More Confidence, Ship with More Control

The best product improvements are rarely the flashiest ones. They’re the quality-of-life changes that make your power users say, “I can’t believe this is here now.”

Our customers use Yoodli to build extremely complex roleplays, think a fourth call with a procurement committee, where the CFO is sharing their screen and critiquing your RFP line by line. Roleplay creation has always been agentic, and our super-admins keep pushing for more flexibility: having one of the AIs leave the call at minute five, asking a learner to share their screen with new pricing mid-conversation.

The more customizable a roleplay gets, the more you want the ability to test it, to see exactly how it will play out before a learner ever steps into it. That’s what Roleplay Suggestions gives you.

How it works

Click “Get suggestions”, and the AI reviews your entire roleplay, checking that the content holds together and that the roles stay exactly where you set them. It’s a way to stress-test a scenario, tighten it, and ship reps that hold up exactly as designed.

Full visibility into complex scenarios

The more parts a roleplay has for context, personas, time limits, files, the title, the more control you have over how real it feels. Suggestions gives you a second layer of review across all of it. Now you can push your scenarios further and still know exactly how everything lines up before you launch.

Precision, down to the role

In a good roleplay, every persona plays its exact part: the AI as the buyer, the learner as the seller, or whatever mix the scenario calls for. Suggestions confirms that alignment holds throughout, so even your most intricate, multi-turn scenarios stay exactly as designed.

Fix it in one click

Every suggestion comes with a plain-language explanation. When the fix is clear, hit Fix. It rewrites only the part that needs changing and leaves your wording, detail, and formatting untouched everywhere else.

More control, not less complexity

This is our first step toward giving admins a genuine second set of eyes on their own work. It’s an AI coach reviewing the roleplay you built, so you can test performance before you ever see it live with learners.