AI Question Generation

An empty cell in the exam blueprint gets filled with original items written for that outcome, and those items go into the institution's own pool. When the mock-exam calendar tightens, the problem facing an assessment team is not finding questions but finding the right one: bought-in sets tend to be items students have already met, or items that sit near an outcome rather than on it. A skills-based curriculum makes that gap harder to hide.

Each item arrives with its difficulty and cognitive level, its distractors built from the mistakes students actually make on that topic, and its answer and worked solution written alongside it. Nothing joins the pool without a similarity check against what is already there and a sign-off from the department, and multiple-choice and open-ended items follow the same route. That removes the need to buy in question sets or quietly recycle last year's mock papers. The pool grows against your own students' data as the term goes on, and writing an exam stops being a job that takes days.