Two pillars. One standard for performance.
— engineered for precision, built for scale
Four stages,
one classification.
Each student step moves through a fixed pipeline before it is labelled — from structured input, through deterministic checks, to model escalation only when needed.
Schema Definition
Each submission is parsed into a structured schema: the mark scheme's criteria, and the student's step-by-step working, validated before any classification begins.
Deterministic Resolution
A rule layer checks each step against the mark scheme's expected forms and a library of known error patterns, resolving clear cases for free with full interpretability.
Escalated Classification
Steps the rule layer can't resolve confidently are batched and passed to a prompted language-model classifier, which uses the full submission to judge intent.
Continuous Fine-Tuning
As labelled submissions accumulate, a smaller open model is fine-tuned on real data, moving the production path toward a faster, self-contained classifier.
Three outcomes,
one taxonomy.
Every step of a student's working is classified into exactly one of three outcomes — a shared vocabulary that ties raw working back to mark-scheme logic.
Correct
The step satisfies the relevant mark-scheme criterion outright — full credit, no ambiguity.
Process Error
The method is right, but execution slipped — an arithmetic mistake, a sign error, a mis-substitution. Method is credited even where a mark is lost.
Knowledge Gap
The method itself is wrong — the step shows the student doesn't understand what this part of the question requires.
System Architecture
Engineered for learning.
F.01
Rule-First Resolution
A deterministic layer checks every step against the mark scheme's expected forms and a library of known error patterns before anything else runs — resolving clear cases for free, with full interpretability.
Explore the rule layerRULE LAYER
LIVEF.02
Model Escalation
Only steps the rule layer can't resolve confidently are escalated to a prompted language-model classifier, which uses the student's complete working to judge whether a step is a procedural slip or a genuine gap in method.
View escalation logicESCALATION
LIVEF.03
Fine-Tuning Roadmap
Once enough labelled submissions exist, a smaller open model is fine-tuned on real classified data, moving the production path toward a faster, self-contained classifier.
See the fine-tuning pathFINE-TUNING
LIVEEarly internal testing on a small validation set showed encouraging results — presented as illustrative preliminary findings, not a verified benchmark.