Domain III · Professional Growth & Development 3 hours Counts towards Level 3 · Certified AI-Enabled Educator Hands-on lab

Module 8 Personalised Learning & Academic Intelligence — the PrepGraph Teacher Workflow

The workflow the whole programme has been building towards, run live in one lab. It is taught first as pedagogy — a way of teaching 40 students as if each were one, which a teacher could do on paper, slowly — and then shown running on PrepGraph with a seeded demo class: the class log carries the lesson to each student’s AI tutor; a short diagnostic from the review-first question bank comes back as results by question and by student; the teacher diagnoses which concept is weak for whom; targeted practice goes out to the students who need it (homework is free for every student) and their AI tutor re-explains it at home (the tutor is part of the student’s AI plan; practicum classes are provisioned with it by the school or pilot); a second check verifies; and the teacher reviews — with early-warning signals beside the results — and decides what happens on Monday. This is not product training. The product is the demonstration; the habit is the point.

CBSE sub-theme 5CBSE sub-theme 6

By the end, teachers can…

  1. 1 Describe the seven-step workflow and say, for each step, who acts — teacher, students or AI — and what decision the teacher makes there.
  2. 2 Record a class log for one lesson (topic, tagged concepts, homework, next test) that a student’s AI tutor can continue at home.
  3. 3 Build an 8–12 question diagnostic from the question bank, assign it to a class, and read the results by question and by student.
  4. 4 Diagnose from a result pattern which concept is weak for which students — and whether it is a prerequisite from an earlier class — before choosing an intervention.
  5. 5 Assign targeted practice to the students who need it and explain to a colleague what the student’s AI tutor does with it at home, and what it does not do.
  6. 6 Run a reassessment on the same competencies and make three review decisions — who improved, who still needs help, what to reteach in class — using results and early-warning signals.

Session plan — 3 hours

TimeActivityFormat
0:00 20 minThe workflow on paper first The seven steps as pedagogy, before any screen: what they would look like with a register, a 10-item test and an evening of marking — and why they rarely happen that way with 40 students. For each step: who acts, what the teacher decides. Nothing about classroom teaching changes; what changes is what becomes possible after the bell. "AI recommends. Teachers decide."Input
0:20 20 minTeach: the class log Each teacher logs today’s demo lesson for the demo class — Class 9 Science, Motion: distance vs displacement, speed vs velocity; homework; next test on Friday — tagging the concepts covered. The facilitator then opens a demo student’s app and shows the tutor picking up the log: it knows what was taught today, and re-explains the homework items the student opens. Two minutes after class, every evening — the hinge of the whole loop.Hands-on
0:40 30 minAssess: build and assign a diagnostic In the question bank, teachers select or AI-draft 8–12 competency questions on the demo lesson’s concepts, check each distractor against a misconception, approve the set (nothing reaches students until they do), and assign it to the demo class as a short diagnostic — not a test. The demo class already carries a seeded set of attempts, so results are waiting when they open the assignment.Hands-on
1:10 30 minIdentify and diagnose Results by question and by student on the screen; the diagnosis sheet on the table. Groups of three work out which concept is weak, for which students, from which distractor — and whether it is really a prerequisite from Class 8. Each group writes a one-line diagnosis and an intervention plan for three named demo students, then shares back. The facilitator presses on every plan: "what did the result say, and what did you decide?"Group work
1:40 25 minIntervene: targeted practice and the tutor at home Teachers assign targeted practice for the students their diagnosis named — homework is free for every student — and share a one-page note to the library so the tutor can answer from it. The facilitator demonstrates the student’s AI tutor on the demo account — by voice in Hindi or English, or by chat in English today — re-explaining tonight’s practice from the class log at the student’s level. Said plainly once: the tutor is part of the student’s AI plan; practicum classes are provisioned with it by the school or pilot; free-tier students still get homework, quizzes, doubts and progress.Hands-on
2:05 20 minVerify: reassess A second short check on the same competencies with different items, assigned from the question bank; the demo class carries seeded reassessment attempts. Teachers put the two result sets side by side, by student, and mark who moved and who did not. Improvement is measured, not assumed.Hands-on
2:25 25 minTeacher review: the Monday decision With the reassessment and the group’s early-warning signals (engagement dropping, scores slipping) open together, groups make three lists — improved, still needs help, reteach in class — and write one teacher-authored line for a parent about one student, using only what the record shows. Then the Monday class log: what gets retaught, what the tutor should carry. This step is the behaviour the certificate certifies.Group work
2:50 10 minClose: the loop in your own class Each teacher names the class, chapter and students they will run the loop with in the practicum, and fills in the first page of the practicum log. One sentence each: "the step I expect to find hardest, and why."Reflection
Total 3 h · timings are a default; facilitators adapt to the group.

Key ideas

A workflow, not a feature

The loop is a way of teaching 40 students as if each were one. A teacher could do it on paper — and some do, slowly, for one topic a term. PrepGraph implements it so it can happen every week. The habit is what the programme certifies; the software is how the habit fits into a timetable.

The class log is the hinge

Two minutes after the bell: topic, concepts, homework, next test. Without it the student’s tutor does not know what was taught today and can only answer in general. With it the lesson continues at home in the teacher’s sequence. Nothing else in the loop works well if this step is skipped.

Diagnose before you intervene

Results say who got what wrong. Only the teacher can say why — this distractor, this concept, a prerequisite from last year, a child who rushed. The wrong diagnosis sends the wrong practice home, and the reassessment will show it. AI recommends; the teacher diagnoses and decides.

Verify, do not assume

A reassessment on the same competencies with different items turns "I think it worked" into "eleven of the twenty-two moved; eleven did not, and that is Monday’s lesson." Measured improvement is what makes the teacher’s review, and later the capstone, honest.

Worked example

9-B, Motion, one week — a demo-class walkthrough (numbers illustrative)

Monday: the teacher logs the lesson — distance vs displacement, speed vs velocity, distance–time graphs; homework, five questions; test Friday. That evening the demo students’ tutors already know it.

Tuesday: a 10-question diagnostic from the question bank; 38 of 40 attempt. Results by question: Q4 and Q7, both graph reading, missed by 22. Q2, displacement with direction, missed by 9 — and the same 9 chose the distractor "distance is always equal to displacement".

Diagnosis: graph reading is a Class 8 coordinate-reading prerequisite, not a physics misconception; the nine have a real displacement misconception. Two different problems, two different interventions.

Wednesday: graph-reading practice for the 22; displacement practice for the 9; a one-page note on reading axes shared to the library. At home, tutors re-explain from the class log — several students ask by voice in Hindi. Friday: a second check, same competencies, different items: graph items now wrong for 11 of 38; displacement wrong for 3 of 9.

Review: early-warning signals flag two students whose engagement dropped — one never attempted either check; that is a conversation, not a worksheet. Monday’s plan: reteach graph reading for the whole class (eleven is still too many), a two-minute check-in with the two, and one teacher-written line to a parent. Illustrative throughout — it is the shape of a week, not a claim about any class.

Assignment — counts towards certification

One full loop on the demo class

Before the practicum launch, run the whole loop yourself on the demo class: log a lesson; build and assign an 8–12 item diagnostic; read the results; write a diagnosis; assign targeted practice to the students who need it; share one note to the library; reassess on the same competencies; write your review decisions. Capture a screenshot or a note at each step.

Deliverable: A 2–3 page loop record: the seven steps in order with one screenshot or note each, a diagnosis paragraph, a review paragraph with the three lists (improved / still needs help / reteach), and one teacher-written parent line for one demo student.

CriterionMeets the standard when…
Loop completeAll seven steps are evidenced, in order, on one class and one chapter — including the class log and the reassessment, the two steps most often skipped.
Diagnosis is reasonedNames the concept (not just the question number), the distractor pattern, and whether it is a prerequisite; says what the pattern does not mean.
Intervention is targetedNamed students get named practice tied to the diagnosis — not "everyone does the worksheet" — and the record says what the tutor would carry home for them.
Review decidesThe three lists are made from the reassessment, not the first check, and at least one classroom action (reteach, check-in, parent line) follows from them.

Where this lands in the PrepGraph workflow

Each certification task corresponds to a behaviour in the teacher portal — so the programme produces a working habit, not a slide deck. The pedagogy comes first; the surface implements it.

Certification taskPedagogical movePrepGraph surface
Teach — log the lesson so the tutor can continue it at homeClass log: topic, concepts, homework, next testClass logRecord what you taught today — topic, tagged concepts, homework, next test — so every student's AI tutor carries the lesson home.
Assess — build and assign an 8–12 item competency diagnosticReview-first items, assigned as a short diagnosticQuestion bank (review-first)Author or AI-draft competency questions; nothing reaches students until a teacher approves it.
Identify — who did not understandResults by question and by studentAssignments & resultsSet a short diagnostic or homework from the question bank; see who attempted, how each question went, and give feedback.
Diagnose — which concept, for whom, prerequisite or misconceptionStudent-level reading of results; teacher decides what it meansStudent profilesOpen one student and see their practice, weak concepts and recent work in one place.
Intervene — targeted practice, carried home and re-explainedPersonalised practice; reteach at the student’s level — voice in Hindi or English, chat in English todayStudent's AI tutorEach student has a personal tutor — voice in Hindi or English, chat in English today — that knows your class log, re-explains the homework items the student opens, and works at their level. Part of the student AI plan — practicum classes are provisioned with it; free-tier students still get homework, quizzes, doubts and progress.
Verify — reassess on the same competenciesMastery check with different itemsAssignments & resultsSet a short diagnostic or homework from the question bank; see who attempted, how each question went, and give feedback.
Teacher review — who improved, who still needs help, what to reteachWeekly review of results with early-warning signalsEarly-warning signalsNightly signals per group — engagement dropping, scores slipping — so you know who needs a check-in this week.

Materials

  • PrepGraph teacher-portal login per teacher, each attached to the same seeded demo class (Class 9 Science, Motion) with diagnostic and reassessment attempts already present
  • One demo student-app login for the facilitator to show the AI tutor (voice in Hindi or English; chat in English today) picking up the class log and re-explaining a homework item
  • Printed seven-step workflow card with the teacher’s decision at each step
  • Diagnosis sheet: question → concept → distractor → students → prerequisite? → intervention
  • Review sheet: improved / still needs help / reteach in class / one parent line
  • Practicum log, first page (the Module 9 template), one per teacher

Facilitator notes

  • Teach the workflow on paper before anyone touches the portal. If a teacher cannot say the seven steps and the decision at each without the product, the lab has not worked.
  • Be plain about the tutor and the plan, once: the tutor is part of the student’s AI plan; the free tier covers homework, quizzes, submissions, doubts and progress; practicum classes are provisioned with it by the school or pilot. Then move on.
  • Do not promise what is not there. There are no automated parent reports and no principal dashboards in this programme; parent updates are teacher-written from what the dashboard shows. Say so if asked.
  • Seed the demo class with diagnostic and reassessment attempts the day before and test every login. If the network fails, run "Identify and diagnose" and "Teacher review" from the printed results sheets — the pedagogy does not need the screen.
  • Run the "tutor picks up the log" step only after the school’s org has the homework-context flag on and verified (the Week-4 readiness gate); otherwise show the class-log pick-up only. Do not show or promise anything the school’s own students will not have the next morning.
  • Every time a teacher reads a result or a signal, ask "what do you decide?" The frame is the same as Module 1: AI recommends, the teacher decides.

Bring the programme to your school

Tell us your board, grades and teacher count. We’ll reply within two working days with a cohort plan and the CPD record templates.