Docs / How it works
How it works.
Ignia turns a goal into a structured course, teaches it in scheduled sessions, and reshapes it from how you actually perform. This section explains that in three passes. First the whole loop, then the part most people want to see up close — how the exercises in front of you were chosen — and then one mechanism worth its own page: what the app does with a wrong idea once it has caught one.
All three pages are written to be checkable. Where a mechanism is deterministic we say so; where something is not yet wired, it is named rather than glossed over.
-
Course engine.
Why Ignia is a course engine and not a chatbot. The three ways your course changes (you edit it, the AI drafts it, the engine schedules it), and the end-to-end loop from your goal to a scheduled session.
Read
-
Exercise sessions.
How a single session is built: the deterministic planner that fixes how many exercises, on which concepts and at which cognitive level, before any AI is called. With real output from the engine, and its honest limits.
Read
-
Misconceptions.
What happens when the app catches you believing something wrong: where the wrong idea is recorded, the three places it is read, the one exercise written to contest it, and exactly what opens and closes it again.
Read
Going deeper.
For exact function names, constants and data structures, the technical section documents the same mechanisms at source level. For what each exercise type tests and how it is graded, see exercise types.