Skip to content

Docs / How it works / Course engine

Not a chatbot. A course engine.

A chatbot answers turn by turn and forgets the shape of what you are learning. Ignia does the opposite: it turns your goal into a structured course: a plan, a concept map, lessons, exercises. It then teaches it in scheduled sessions and reshapes it from how you actually do. Every change to your course has a traceable cause: something you edited, your goal, or your performance.

Three ways your course changes.

The same piece of your course, say a lesson, can change through more than one of these. They are different kinds of change, with different guarantees.

You shape it

Direct edits.

Edit the brief, the plan, a lesson, the concept map, your sources, the schedule. Your version is authoritative: the app keeps it, and folds it into what it generates next.

The AI drafts

Generation & revision.

From your goal and your sources, the AI writes the plan, the lessons, the concept map and the exercises. Ask for a revision and it re-drafts in place, keeping what already works instead of starting over.

The engine adapts

Scheduling, no guesswork.

A planner turns your plan into sessions. A spaced-repetition engine tracks what you are retaining and brings a concept back before you forget it. That part is math, not an AI hunch.

The loop, end to end.

Your goal seeds a course; the course becomes sessions; your answers feed an engine that decides what comes back and when.

  1. 1

    Your goal → a plan. You describe what you want to learn. The AI reads it (plus any sources you add) and drafts a course plan and a map of the concepts and how they depend on each other.

  2. 2

    Plan → sessions. A deterministic planner, not the AI, slices the plan into scheduled sessions on the days you picked, and lists which concepts each one teaches.

  3. 3

    Session → exercises. For each session the engine picks the right concepts (new material plus reviews that are coming due) and the AI writes exercises tuned to where you are on each one.

  4. 4

    Your answers → adaptation. How you do updates a retention model per concept. That feeds three things: what future sessions review, when a concept is due again, and which lessons the app offers to revise for you.

You stay in charge.

Nothing is a black box you cannot touch. You can edit the plan and lessons by hand, ask for a targeted revision, correct the concept map, or change a grade, and the app treats your correction as the truth. Where the app adapts on its own (scheduling reviews), it does so from your performance, not from an opaque model of you.

Under the hood.

Want the complete map: every object, every mechanism, and exactly what context each generation reads? See the technical documentation.