Generic mocks can't close individual gaps. CGF captures learning signals from every evaluation and powers test personalization at coaching scale.
Your coaching platform runs hundreds of mocks per month. Evaluators mark them, scores come back. You can see who's performing and who isn't.
But your academic team can't tell you whether the student who scored 58% is weak on organic chemistry or physical chemistry. That knowledge has to be earned through evaluation infrastructure that captures signals — not just scores.
All students sit the same paper. High performers waste time on mastered questions. Struggling students hit questions beyond their level. Neither benefits.
Your evaluation stack returns a score. Without concept-level signals, you have no basis for building a personalized test.
Students fail the same question types test after test. Without gap detection, the platform keeps exposing the same weakness without fixing it.
Personalization requires concept-level signals as input. Score-only evaluation can't produce the gap map personalized test generation requires.
Coaching platforms trying to personalize testing face the same ceiling: they have content, but they don't have the learner intelligence to route it correctly. Adding more questions doesn't solve personalization. Building the memory layer that knows each student's gap does.
Personalized testing is an output. The input is a learning memory built from evaluation signals captured over every mock and every assessed response. The richer that memory, the more targeted the test. The more targeted the gap closure.
SAME CONTENT LIBRARY. COMPLETELY DIFFERENT TEST.
Every evaluation event feeds the gap map. Every gap map update sharpens the next test.
Signal Capture
Every evaluated response is logged with subject, concept, score, and timestamp.
Gap Map Update
CGF updates each learner's concept-level gap map in real time as evaluations complete.
Personalized Test Signal Output
A test_recommendation object is generated: focus concepts, question weight, difficulty setting.
Continuous Loop
Every new test response feeds back into the gap map — every gap map update contributes to the evaluation infrastructure that powers personalisation at scale.
// personalized_test_signal.json
{
"learner_id": "JEE_2026_0847",
"gap_map": {
"organic_chemistry": {
"score_avg": 0.42,
"trend": "DECLINING",
"recurrence": 4
},
"physical_chemistry": {
"score_avg": 0.61,
"trend": "IMPROVING",
"recurrence": 2
},
"inorganic_chemistry": {
"score_avg": 0.84,
"trend": "STABLE",
"recurrence": 1
}
},
"test_recommendation": {
"focus": ["organic_chemistry", "reaction_mechanisms"],
"weight": { "focus": "70%", "review": "30%" },
"difficulty": "ADAPTIVE"
}
}Personalisation is only as good as the evaluation it runs on. These numbers are from live deployments.
93%
Scoring Reliability (ICC)
Scores you personalise from are scores you can trust.
87%
Responses Within Tolerance (±6 marks)
Gap map built on consistent signal, not noise.
90%+
Accurate Feedback Delivered Instantly
Every test response generates a signal. No manual review queue.
Human-in-the-Loop by Design
Every personalised test path includes a teacher control point. Educators review gap maps, override recommendations, and validate learning signals before they become curriculum decisions.
Personalisation is a teacher-controlled process. Every recommendation is bounded, validated, and reversible.
Teacher Control Point
Educators review and override any gap map entry. A learner's concept weakness is not locked in — teachers validate, adjust, or remove signal data before it shapes test delivery.
Signal Validation Before Recommendation
CGF does not generate test recommendations from raw scores. Every signal passes through the evaluation reliability layer — 93% ICC alignment — before it enters the gap map.
Curriculum-Bounded Generation
Personalisation operates within the syllabus scope defined by the institution. No question, concept weight, or difficulty setting falls outside the mapped curriculum boundary.
Not feature claims — results observed in live deployments where evaluation feeds personalisation.
Recurring Gaps Close
Students stop failing the same concepts repeatedly. The gap map detects recurrence and increases concept weight until the signal clears.
Test Time Becomes Targeted
Learners spend test time on concepts where signal says they need it. Every minute of assessment generates a signal that updates the next test.
Teacher Visibility Improves
Educators gain a concept-level view of every learner — not just scores, but gap trends, recurrence patterns, and signal history across the term.
Learning Progression Is Measurable
Gap map trend lines show whether an intervention is working. Institutions can track concept mastery across cohorts, not just average scores.
Personalised testing is the middle layer — built on evaluation signals, feeding into structured learning paths.
Exam Evaluation at Scale
Evaluation at exam scale feeds the gap maps that personalised testing draws from. The more evaluations run, the sharper the signal.
See use case →Automated Subjective Grading
Personalised test signals depend on consistent subjective grading. Automated subjective grading is the reliability layer underneath the gap map.
See how it works →Action Plans
Gap maps don't stop at test generation. Action plans convert signal patterns into structured learning paths — the next step after personalised testing.
See the product →Personalised testing looks different depending on your platform. Here's where each segment fits.
Students repeat mocks. Gaps don't close.
Personalised test signals built on evaluation signals replace mock repetition with targeted concept work.
See coaching use case →Adaptive learning with no evaluation layer underneath.
CGF connects your existing content library to evaluation signals — personalisation with structured learning intelligence.
See EdTech use case →Assessment data sits in the LMS. It never feeds the learner's next test.
CGF bridges evaluation output and test generation — your LMS data becomes a live signal layer.
See LMS use case →One syllabus. Wildly different learner starting points.
Personalised testing adapts within your curriculum boundary — no learner gets the same test twice without a signal reason.
See institutions use case →// COMMON.QUESTIONS
Personalised testing in education is a system where each learner's next test is shaped by their concept-level gap map — built from prior evaluation signals. CGF captures those signals during evaluation, structures them into a gap map, and generates test recommendations that adapt to each learner's current state.
What is personalised testing in education?
Personalised testing means each learner receives a test shaped by their individual concept gaps — not a shared question bank served uniformly. CGF builds that gap map from evaluation signals: every assessed response updates the learner's concept-level state, and the next test recommendation draws directly from it.
How does CGF generate a personalised test signal?
CGF evaluates each response, captures concept-level signals, and updates a learner's gap map in real time. From that gap map, it generates a test_recommendation object specifying focus concepts, question weight distribution, and difficulty setting — without requiring a teacher to manually review every learner's history.
Can teachers override personalised test recommendations?
Yes. Every gap map entry is editable by educators. Teachers can review concept-level signals, adjust weakness classifications, and override test recommendations before they reach the learner. Personalisation in CGF is a teacher-controlled process — not a black-box algorithm.
What proof numbers support CGF's personalised testing accuracy?
CGF achieves 93% reliability in total scoring alignment (ICC), with 87% of responses within tolerance of ±6 marks. Feedback accuracy exceeds 90% and is delivered instantly. These numbers reflect the evaluation reliability layer that all personalised test signals are built on.
How is CGF different from standard adaptive testing platforms?
Standard adaptive testing adjusts question difficulty based on right/wrong signals alone. CGF builds a concept-level gap map from full subjective evaluation — capturing what the learner understands, where they err, and how that pattern recurs. The test signal is richer, and the personalisation is grounded in reliable evaluation signals.
// RELATED · RESEARCH & PLAYBOOKS
Book a demo and see how CGF turns evaluation signals into a personalised test system — built for your platform, governed by your teachers.