Branching scenarios for behavior change
Three practice conversations in two formats. In the pharmacy scenarios you talk freely, and the character only tells you what your questions earn. In the classroom scenario you move through a fixed branching tree. Every debrief scores what you did, not whether you reached the last screen.
Design demonstration. Not yet approved for learner use.
Success means a change in what people do after the scenario, not finishing it.
The design standard for both scenarios, from the research guide behind this app.
Healthcare · pharmacy leadership coaching
Open conversation · fact packet
You're Marcus. Coach a technician after a pattern of labeling errors. Open conversation, no scripted choices.
You play: Marcus Delgado. Execution. He tells instead of asks. You talk with: Sam, pharmacy technician at gilbert's.
Healthcare · pharmacy leadership coaching
Open conversation · fact packet
You're Priya, a new manager. A technician has been rushing customers. Find out why before you decide what it means.
You play: Priya Nair. Conceptual. She hasn't built the mental model that asking is the work. You talk with: Dev, pharmacy technician at gilbert's.
Education · teacher professional development
Fixed branching tree
A student's line-following robot broke before the qualifying run. Help them find the cause without taking over, while another student waits.
You play: Classroom teacher You talk with: Jordan, a student whose micro:bit line-following robot stopped working before the qualifying run.
Each scenario starts from a behavior an observer could see, not a topic.
Duration, the moment the skill breaks down, and learner experience set the scope.
Characters follow written response rules. Warmth alone doesn't unlock evidence.
Open conversations built on a fact packet, next to a fixed branching tree. Same guardrails, different trade-offs.
Mistakes play out, and you can repair them. Repairs stay on the record.
The AI labels what you did on each turn. Code turns those labels into outcomes and scores, so a wrong label means a wrong score. Every label is shown and can be flagged.
An authoring kit turns one filled-in template into a full practice conversation, then runs the same rule checks the site runs before release.
Personal information, off-topic chat, personal and clinical advice, and rule overrides are caught.