What AI unlocks · adaptive assessment

A quiz measures recall. This one measures judgment.

Multiple choice can't tell you whether someone would actually handle customer data safely on a Tuesday afternoon. An interviewer can. This check asks open questions, scores the substance of the answer, and spends its final question wherever you looked weakest. What lands in the LRS is a mastery profile per competency.

🧭

Skill Check

Customer data handling · adaptive

Guided demo

The setup

A quiz asks ten questions. An assessor asks the right five.

This is a knowledge check on customer data handling that behaves like a good interviewer: one open question per competency, scored on substance, then the final question spent wherever you looked weakest.

Every exchange emits xAPI, so the record isn't "completed the quiz." It's a mastery profile per competency, the thing the manager dashboard rolls up.

Running as a guided version: pick from realistic answers, the same scoring and adaptivity applies.

Open answers, real signal

Recognition is cheap; recall under a realistic scenario is the skill. Free-text answers scored on substance separate "picked the right option" from "would do the right thing."

Adaptive by design

Fixed quizzes spend questions evenly, on strengths and weaknesses alike. This one watches the mastery bars and puts its last question where the risk is.

Feeds the dashboard

Each answer is an xAPI statement with a scaled score per competency. That's the learner-side instrument behind the manager dashboard: gaps by theme, not completion rates.

When a model is connected you answer in your own words and it scores what you actually said; otherwise it runs as a guided version off the same design, fully playable either way.