Designing an AI tutor that supports — rather than replaces — the learner’s judgment

In MED.Z, Zaiko uses a student’s own schedule and exam-performance history to suggest what to study next. The learner can see the evidence behind the recommendation and remains responsible for the final plan.

We’re exploring the right balance between automation, transparency and control for a high-stakes education context. I’d value examples of products that earned trust through explainable recommendations, good defaults and clear override mechanisms.

Product: medz.app.br/