Anyone can generate questions. The hard part is telling you what your answers mean.
“These questions are nothing like the exam”
Exam mode uses the real question count and the real clock. No feedback until you submit, and the timer keeps running if you walk away, because that is what happens on the day.
Exam mode
“So what do I study next?”
Not "revise Model the data". More like: you missed TREATAS, CROSSJOIN and ALLEXCEPT in one session, so filter context is the thing to sit down with, and here are 20 questions on it.
AI study plan
“I'm improving… I think?”
Not just your last attempt. Where each topic started, where it is now, whether you're rushing, and which improvements are real rather than one lucky set.
Topic trends