How it works

How Annelia
thinks.

No magic. A few well-chosen algorithms doing their jobs.

§1 · The mastery model

A probability, not a percentage.

Each student × concept carries its own evolving distribution.

Beta(α, β) playgroundReal product UI lands here in P3.

Posterior

p(mastery | evidence) ∝ Beta(α + s, β + f)

s = positive signal weight     f = negative signal weight

Signal weight is bounded — a single answer can't claim more
certainty than the system can earn.

§2 · Post-turn pipeline

Five stages. Always async.

Never blocks the next message.

  1. 01

    Concept extraction

    What did the student engage with? Not what's in the document — what got touched.

  2. 02

    Confidence classifier

    Confident, hesitant, or confused? Tone is signal.

  3. 03

    Bayesian update

    Posterior shifts by evidence weight. Self-explanation > a passive okay.

  4. 04

    Misconception check

    Conservative — only flags above 0.6 confidence. False positives erode trust.

  5. 05

    Next-question generation

    Pre-generates the next check, contextual to actual gaps.

§3 · Prerequisite graph

Concepts know what they need.

Auto-extracted. Teacher-overridable.

Concept dependency walkerReal product UI lands here in P3.

§4 · Intent-aware retrieval

Different questions.
Different strategies.

Definition, derivation, exercise — each routed differently.

Question → intent → BM25 + dense → rerank → top-KReal product UI lands here in P3.

§5 · The boundary

Stays inside
what you teach.

Enforced at the database, not the application.

Architectural diagram — content core, AI orbit, hard boundaryReal product UI lands here in P3.

§6 · Misconception detector

Wrong answers and wrong models.
Different things.

Conservative threshold. Teacher-trust comes first.

Class-wide misconception rankingReal product UI lands here in P3.

§7 · The operator layer

The school's controls. Exposed.

  • Cost rollup

    Per-school, per-event-type, per-time-window.

  • Model dispatcher

    Swap which model handles which task. No redeploy. Propagates in 60s.

  • Audit log

    Every LLM call, embedding, retrieval — timestamped, costed, attributable.

  • Approval gates

    Teacher onboarding, document upload, configuration changes.

See it on your school's data.

Talk to us →