How the roadmap engine works

A static study timetable assumes you know exactly what you need to work on and exactly how long it'll take. That's almost never true. That's why we built a dynamic roadmap engine instead.

When you create your roadmap, Claritii starts by loading your subjects, assessments, and schedule blocks. Then it runs a proportional-sampling algorithm that distributes sessions across subjects weighted by urgency - how close each assessment is, how weak you are on those topics, and how much time you've blocked out.

Each session gets a content type (flashcards, quiz, practice paper, or summary) based on a rotating schedule that ensures variety. You don't get three flashcards sessions in a row - the engine mixes them so you're constantly switching between recall, application, and review.

As you complete sessions, your mastery scores update (using an SM-2-inspired Bayesian variant - think spaced repetition on steroids). The next time the roadmap regenerates, it factors in your updated mastery, any new assessment dates, and your actual study pace. If you're breezing through a topic, it'll schedule less time. If you're stuck, it'll add more sessions.

The roadmap also pre-generates content for the next 3 days of sessions. That means when you open the app, your flashcards, quiz questions, and practice papers are already there - no loading spinner, no 'AI is thinking...', just study.

We're continually refining the engine. The current version scores roadmaps on five dimensions: feasibility, workload balance, revision quality, deadline coverage, and spaced repetition quality. We use these scores to diagnose weak roadmaps and suggest improvements.