How it works

From a test result to a verified next step.

Heramb.study is designed around one repeating loop. Here's what happens at each step, a worked example using fictional data, and what each step can't do.

  1. Step 1

    Assess

    Every question is tagged to a topic and skill, so a score becomes a map of what each learner can and can't yet do.

  2. Step 2

    Diagnose

    Marks are split by topic and by the kind of mistake — a concept gap needs different help from a careless slip or time pressure.

  3. Step 3

    Practise

    Each learner gets a small set of questions aimed at their weakest topics instead of repeating a whole chapter.

  4. Step 4

    Verify

    A follow-up check on the same topics shows whether the gap actually closed — not just whether practice was completed.

  5. Step 5

    Improve

    Teachers decide what to reteach, who to group together, and what to share with students and parents, based on verified change.

Then the loop repeats with the next topic.

Each step in detail

  1. Step 1

    Assess

    Every question is tagged to a topic and skill, so a score becomes a map of what each learner can and can't yet do.

    Example
    A 30-question Physics unit test, where each question is tagged to a topic (e.g. Rotational motion) and a skill (e.g. applying torque).
    Limitation
    A test can only reveal what it asks. Topics with one or two questions give weak evidence; we'll flag these rather than over-interpret them.

  2. Step 2

    Diagnose

    Marks are split by topic and by the kind of mistake — a concept gap needs different help from a careless slip or time pressure.

    Example
    Aarav (fictional) scores 41% on Rotational motion. Most wrong answers use the wrong torque direction — a concept issue, not carelessness.
    Limitation
    Mistake types are inferred from answer patterns and teacher input. They are suggestions for a teacher to confirm, not automatic verdicts.

  3. Step 3

    Practise

    Each learner gets a small set of questions aimed at their weakest topics instead of repeating a whole chapter.

    Example
    Twelve questions on rotational motion, ordered from a worked example to independent problems. About 25 minutes.
    Limitation
    Practice helps only if it's done and if it targets the real cause. Some gaps need a teacher's explanation before any practice.

  4. Step 4

    Verify

    A follow-up check on the same topics shows whether the gap actually closed — not just whether practice was completed.

    Example
    Three weeks later, a short check on the same topic at similar difficulty: 63%. A real, verified change — but not yet secure.
    Limitation
    Two tests are never identical. Small changes are reported as 'no clear change' rather than as progress.

  5. Step 5

    Improve

    Teachers decide what to reteach, who to group together, and what to share with students and parents, based on verified change.

    Example
    The teacher keeps Aarav on rotation practice, groups him with two others with the same gap, and shares a clear report home.
    Limitation
    The decision stays with the teacher. Heramb.study informs judgement; it doesn't replace it.

The result: a report with a next step

What a teacher, student or parent would see after one full cycle.

Progress report — Aarav (fictional student)

Class 11 · Physics · Evening batch (fictional)

Illustrative — fictional data

Unit test 3

65%

Follow-up check

74%

Biggest verified change

Rotational motion +22 pts

Topic results before and after targeted practice (fictional)
TopicBeforeAfterStatus nowMain mistake type
Kinematics82%85% (+3)SecureCareless
Laws of motion68%77% (+9)SecureProcedure
Work, energy & power59%71% (+12)DevelopingProcedure
Rotational motion41%63% (+22)DevelopingConcept
Gravitation74%72% (−2)DevelopingTime pressure

Suggested next step: continue short practice on rotational motion (concept: direction of torque). Gravitation slipped slightly — the mistakes were under time pressure, so a timed mini-set is more useful than re-teaching.

Thresholds used: secure ≥ 75%, developing 50–74%, learning gap < 50%. Two tests are only comparable when they cover the same topics at similar difficulty.

You can produce a report like this by hand today with our free progress report template.

Invite-only early access

Help shape Heramb.study from the first version.

We're opening access in small, invite-only groups so we can learn from real classrooms before scaling. Requesting an invite is free and doesn't commit you to anything.

Request an inviteExplore how it works →