01
Map
Maintain the approved curriculum, concepts, prerequisites, and academic plan.
How it works
From everyday evidence to a living model of learning.
Cuegence maintains two foundational knowledge structures—the approved curriculum model and the longitudinal learner model—and an evidence loop between them. This is how the loop runs.
The learning-intelligence loop
Every useful signal makes the next decision more informed.
Cuegence fits around normal academic activity. It continuously connects curriculum, classroom evidence, teacher judgment, and subsequent outcomes.
02
Teach
Prepare with class context. The teacher teaches normally and confirms coverage.
03
Measure
Ingest question-wise marks, assignments, homework, and teacher feedback.
04
Model
Update concept-level understanding while preserving provenance and uncertainty.
05
Act
Guide teaching, interventions, home practice, and leadership attention.
06
Learn
Use subsequent evidence to revise the interpretation and improve the next decision.
Closed loop, not one-way generation
The system learns whether an intervention helped from what happens next.
Model at the centre
Two models. One continuously improving view of learning.
The curriculum model describes what should be learned. The longitudinal model describes what the evidence currently suggests has been learned. Cuegence becomes more useful where they meet.
Product test
A feature should improve the learning model with useful evidence—or use that model to make a better academic decision.
Model A
Curriculum / concept model
Board / curriculumGrade & subjectLearning outcomesAcademic plan
Model B
Longitudinal learner model
Concept evidence history
Pseudonymous learner · L-0714
12 Aug
Assessment
Potential gap
14 Aug
Feedback
Confirmed
17 Aug
Practice
Improving
20 Aug
Recall
Retained
Provenance retained
Confidence calibrated
Teacher-correctable
Evidence without extra burden
Intelligence built from work schools already do.
Cuegence prioritizes academic evidence already produced through normal school workflows. Every signal keeps its source, context, and time—so interpretation remains inspectable.
Target: ≤5 minutes of Cuegence-specific administration per class each day
Question-wise marks
Tests, unit assessments, and examinations
Concept mapping
Questions mapped to concepts and learning outcomes
Assignments & homework
Issued in Cuegence or imported from school systems
Teacher feedback
Corrections, observations, and confirmed misconceptions
Curriculum coverage
What was actually taught and when
Existing systems
Question-wise marks entry via Excel and API import
AI-assisted evaluation
Permitted student work, with PII excluded and teacher review rules
Future approved signals
Structured oral evidence, classroom checks, and external assessments
Evidence
Source retained
Nothing is silently overwritten.
Interpretation
Probabilistic and revisable
Conflicting evidence changes confidence.
Teacher judgment
High-weight and auditable
Corrections update, never erase, history.
Evidence principles
Interpretation you can always trace back.
The quality of Cuegence depends on the density, diversity, provenance, and usefulness of the signals feeding the longitudinal model. These rules govern how every signal is treated.
01
Provenance always retained
Every signal keeps its source, timestamp, context, and mapping to relevant concepts wherever possible.
02
Inference never overwrites evidence
Interpretations are derived from evidence and can change; the underlying evidence itself is never replaced.
03
Conflicting evidence is expected
Contradictory signals affect confidence rather than being silently discarded.
04
Teacher correction is high-weight
A teacher’s correction carries strong weight, but remains auditable rather than destructively replacing history.
05
Absence of evidence is not evidence of absence
The system distinguishes “we have no signal yet” from “the evidence suggests poor understanding.”
06
Confidence is calibrated empirically
Evidence weighting and confidence calibration are core model-design problems, evaluated against real outcomes.
Pilot programme
See the loop run on your school's own evidence.
Pilot schools evaluate model-informed teaching against curriculum-only output in real academic workflows.