Programme B · 3 hours · school leaders

Leading an AI-Enabled School

A three-hour working session for the people who decide whether AI in the school becomes a teaching habit or a purchase that gathers dust.

3

hours, one morning

Week 0

of the 12-week plan

4

deliverables

1

policy draft

For principals and vice-principals · academic coordinators and heads of department · trustees and management representatives · the person who will own the school’s ai policy

Why leadership first

Teacher adoption dies without ownership

The teacher programme creates the habit; this session creates the conditions for the habit to survive a school year. That is why it runs before any teacher sits down.

Adoption dies without ownership

Teachers protect their time, rightly. If management has not blocked the sessions, the practicum weeks and the weekly clinic on the annual plan, Week 5 collapses under unit tests, events and substitutions.

Policy before practice

Students and parents will ask teachers what the school allows. Without a stated position, every teacher improvises — and the strictest and loosest answers both end up in the PTM.

Evidence needs a reader

The programme measures activation, not attendance. If nobody in leadership reads the funnel each month and asks a teacher “what did the diagnostic show?”, the programme ends at the certificate.

Decisions only leaders can make

Which teachers, which sections, how CPD hours are recorded, who owns the policy, who the AI Champions will be — none of this can be decided in a staff room.

Week 0 · Leadership

Principal & management orientation

The 3-hour leadership workshop: what AI changes, what it must not replace, policy, data, the 12-week plan. Decide the cohort and the CPD recording approach.

Week 0 of the 12-week plan, on campus, before the teacher cohort begins. Runs as one morning or two 90-minute sittings. Eight to twelve people round one table works best; a trust can bring its schools together.

The full 12-week plan

The first hour

What AI changes in a school — and what it must not replace

AI recommends. Teachers decide. Leaders need both lists by heart, because staff and parents will ask for them.

What AI changes

  • Visibility: per-concept, per-student, refreshed nightly — not a mark at the end of term.
  • The first pass: who did not understand, and which question, before the teacher has marked a single script.
  • Practice at each student’s level — by voice in Hindi or English, by chat in English today — carried home after the lesson.
  • Drafting: competency questions, worksheets, explanations at three levels, parent notes — reviewed by the teacher before use.

The AI: analyses · suggests · personalises · practises · tracks · assists

What it must not replace

  • The teacher’s judgement about a child — a pattern in data is a reason to look, not a verdict.
  • The classroom relationship: motivation, challenge, context, trust.
  • Accountability for anything that reaches a student or a parent.
  • Teaching itself — the workflow begins with “teach the chapter as you always have”.
  • Decisions about promotion, reports, placements or discipline.

The teacher: diagnoses · guides · challenges · contextualises · motivates · validates

By the end, the leadership team can…

  1. 1 Explain to staff and parents, in two minutes, what AI changes in teaching and name three things it must not replace.
  2. 2 Produce a one-page draft of the school’s AI and responsible-use policy with a named owner and a review date.
  3. 3 State the school’s position on student AI use and academic integrity in one sentence a Class 9 student can repeat.
  4. 4 Ask any vendor — including PrepGraph — the seven data questions and recognise an incomplete answer.
  5. 5 Describe the teach → assess → identify → diagnose → intervene → verify → review workflow and the teacher/AI split in their own words.
  6. 6 Read the activation funnel for a cohort and name the leadership move that takes teachers to the next stage.
  7. 7 Choose process and evidence measures for the year that do not depend on anyone promising score gains.
  8. 8 Fix the 12-week calendar: cohort, CPD recording approach, policy owner and AI Champion nominations, each with a name and a date.

Session plan — 3 hours

A working session, not a presentation. Two of the eleven blocks are group work on the school’s own documents; the drafts leave the room.

TimeActivityFormat
0:00 10 minOpening: what is already happening in your school Round the table: where AI showed up in the school this month — a teacher drafting with a chat assistant, students using one for homework, a parent asking at the gate. Captured on one sheet; it becomes the “current state” line of the policy.Discussion
0:10 20 minWhat AI changes in education — and what it must not replace Prediction versus understanding, in five minutes. Then the split: AI analyses, suggests, personalises, practises, tracks, assists; the teacher diagnoses, guides, challenges, contextualises, motivates, validates. Three short cases where a school would have gone wrong by letting a dashboard decide.Input
0:30 30 minGovernance: draft the school AI & responsible-use policy Two groups, one-page template: purpose; permitted and prohibited uses for staff; the same for students; data and privacy; academic integrity; owner and review cycle. Groups swap drafts and mark what is missing. The draft leaves the room with the principal.Group work
1:00 15 minStudent AI policy and academic integrity Using AI to learn versus using it to avoid learning. Three real homework tasks from Classes 8–10: which survive a chat assistant unchanged, which need redesign. The school writes what counts as a student’s own work, in one sentence.Discussion
1:15 20 minData and privacy Children’s data under the DPDP Act in plain language; who the data fiduciary is. The seven questions to ask any vendor — PrepGraph included — and what a complete answer looks like. What PrepGraph holds and why, read from the list on this page.Input
1:35 15 minTeacher adoption and the activation funnel Why “trained” is the weakest number in the funnel. The nine-stage funnel; for each stage, the one leadership move that gets teachers to the next — protected time, a timetabled clinic, recognition in the staff meeting, a principal who asks about the diagnostic rather than the attendance.Discussion
1:50 20 minThe seven-step workflow, with forty students Walked through on Class 9 Science — Motion: teach and log it; a ten-question diagnostic; who did not understand; which concept; named practice carried home by the student’s AI tutor; a reassessment; the teacher’s review. Leaders see what a teacher will actually do in Weeks 5–8 and what to ask them about.Demonstration
2:10 15 minMeasuring academic outcomes honestly Process metrics first; coverage second; the school’s own before/after per chapter third. Why no vendor should promise score gains in a term, and what to say when one does. One illustrative capstone result, labelled as illustrative.Input
2:25 10 minParent communication Teacher-written, AI-assisted updates drafted from dashboard data and read by the teacher before sending; the 45-minute parent AI-awareness session at the next PTM; how to explain the student AI tutor and the AI plan in one paragraph.Discussion
2:35 20 minThe 12-week roadmap and the decisions in the room Week 0 to Week 12 laid over the school’s own calendar — exams, events, staff-development days. Then the decisions: cohort, CPD recording approach, policy owner, AI Champion nominations. Each written with a name and a date.Group work
2:55 5 minClose: commitments Each leader writes one thing they will say to staff on Monday and one thing they will stop doing. The four deliverables are checked off before anyone leaves.Reflection
Total 3 h · timings are a default; facilitators adapt to the group.

Governance, data and privacy

A policy with an owner, and questions for every vendor

The school AI and responsible-use policy is drafted in the room from a one-page template. The data questions are the same for every system the school uses — PrepGraph answers them too.

The school AI & responsible-use policy

One page: purpose; permitted and prohibited uses for staff; the same for students; data and privacy; academic integrity; a named owner and a review cycle. Drafted in the governance block, checked against the student-policy and data blocks, and taken away by the principal.

Open the policy template

Children’s data — the context

India’s Digital Personal Data Protection Act, 2023 treats children’s data with particular care: processing a child’s data needs verifiable parental consent, and tracking, behavioural monitoring or targeted advertising directed at children is not permitted.

The school should know who the data fiduciary is for each system it uses, hold the consent records, and be able to name the person a parent contacts with a question. The school’s counsel decides the detail; this session makes sure the questions are asked.

How this applies to learning analytics a school itself commissions — per-concept strength estimates, the nightly early-warning flags a teacher sees for a group — is for the school’s counsel and the school agreement. It is one of the seven questions below, and PrepGraph answers it in writing like any other vendor.

Seven questions to ask any vendor

  1. 1 What student data do you collect, and for what stated purpose? Is anything collected that the purpose does not need?
  2. 2 Who is the data fiduciary for student data — the school or you? What does the agreement say?
  3. 3 How is verifiable parental consent obtained and recorded, and how can a parent withdraw it?
  4. 4 Where is the data stored, who can access it, and is it ever shared with third parties or used for advertising?
  5. 5 How long is data kept, and how does the school have it deleted when the agreement ends?
  6. 6 If there is a breach, who is told, and within what time?
  7. 7 Can a teacher see what the AI told a student? Who has the last word when the AI is wrong?

What PrepGraph holds

Roster data the school provides
Student name, class and section, and a guardian contact used for enrolment and consent.
What teachers create
Class logs, assignments and questions, documents shared to the library, replies to doubts.
What students do on the platform
Assignment attempts and results, doubts asked, and — for students on the AI plan — their AI-tutor conversations.
Derived signals
Per-concept strength estimates and the nightly early-warning flags teachers see for their groups.

Ask us the seven questions above. Our written answers on purpose, consent, access, retention, deletion and breach handling form part of the school agreement — the same standard the policy sets for every other vendor. Students linked to a school get the free basics. The AI plan is either provisioned by the school or pilot, or bought by a parent with the school’s consent — the school’s arrangement decides which. PrepGraph does not market plans to a school’s students without that arrangement.

Teacher adoption

The activation funnel — and the leadership move at each stage

Success is measured as activation, not attendance. Each stage has one thing only leadership can do.

  1. 1 Trained

    Put the sessions on the annual plan and protect them; attend the first and last yourself.

  2. 2 Logged into PrepGraph

    Make sure every cohort teacher has an account and a working device in Week 1 — not Week 4.

  3. 3 Created or selected a class

    Decide the practicum sections with the teachers, not for them; one class, one chapter, 10–30 students.

  4. 4 Reviewed a student diagnostic

    In Week 5, ask one question in the staff meeting: “What did your diagnostic show?”

  5. 5 Identified a learning gap

    Ask for the concept, not the mark: “Which concept was weak — and was it a prerequisite from last year?”

  6. 6 Assigned an intervention

    Protect the weekly 30-minute clinic; that is where interventions get assigned.

  7. 7 Reviewed student progress

    Read the funnel monthly with the coordinator; notice teachers stuck at a stage and ask what is in the way.

  8. 8 Used the data in a classroom decision

    Recognise a decision made on evidence — a reteach, a parent note, a regrouping — in front of colleagues.

  9. 9 Repeated the workflow

    Make the second turn of the loop the norm for next term’s chapters; nominate AI Champions who will carry it.

AI-enabled personalised learning

The seven-step workflow your teachers will run

Walked through in the session on Class 9 Science — Motion. Leaders do not need to operate it; they need to know what to ask a teacher about it in Week 6.

  1. 1

    Teach

    The teacher teaches the chapter as they always have — and records what was covered.

    Nothing about classroom teaching changes. The class log (topic, concepts, homework, next test) is what lets the AI tutor continue the lesson at home.

    Teacher

  2. 2

    Assess

    Students complete a short diagnostic — 8 to 12 competency questions, not a test.

    Questions are written (or AI-drafted and teacher-approved) against the concepts of the chapter, with distractors that catch specific misconceptions.

    Students

  3. 3

    Identify

    Who did not understand?

    Results by student and by question — the first pass that replaces the spreadsheet.

    AI

  4. 4

    Diagnose

    Exactly which concept is weak — and is it really a prerequisite from last year?

    The teacher reads the pattern (which distractor, which concept, which students) and decides what it means. AI recommends; the teacher decides.

    AI + Teacher

  5. 5

    Intervene

    Named students get named practice — and their AI tutor carries it home.

    The teacher assigns differentiated practice or a reteach for the nine students who need it; the student’s tutor re-explains at their level — by voice in Hindi or English, by chat in English today.

    AI + Teacher

  6. 6

    Verify

    Students are reassessed on the same competencies.

    A short second check — same concepts, different items — so that improvement is measured, not assumed.

    Students

  7. 7

    Teacher review

    Who improved? Who still needs help? What needs reteaching in class?

    The teacher closes the loop: a classroom decision, a note to a parent, a reteach on Monday. This is the behaviour the certificate actually certifies.

    Teacher

Plainly: the student’s AI tutor is part of the student’s AI plan. In the institutional programme the practicum classes are provisioned with it by the school or pilot; every student, on any plan, gets homework, quizzes, doubts and progress. The workflow is a pedagogy; PrepGraph is how it runs with forty students.

Measuring academic outcomes honestly

Process first. Then coverage. Then the school’s own evidence.

No vendor should promise score gains in a term. The school measures what it can see and own.

1. Process metrics first

The activation funnel: trained → logged in → class set up → diagnostic reviewed → gap identified → intervention assigned → progress reviewed → decision made → repeated. Read monthly. It tells you whether the workflow is becoming a habit before any academic number could.

2. Coverage second

Of the practicum chapters, how many had a diagnostic and a reassessment? What share of students in those sections were reassessed? How many teacher-written parent notes went out? These are the school’s own numbers and need no vendor.

3. The school’s own before/after, third

Per teacher, per chapter: mastery on the same competencies before and after the intervention, and how many students still need remediation. This is the teacher’s evidence — it belongs in the capstone and in the school’s records, labelled as what it is.

What not to accept

No vendor should promise score gains in a term. Board-exam movement is slow and hard to attribute; a promise to “boost results” is a reason to ask the seven data questions more carefully. PrepGraph makes no such promise. We show the workflow and the teacher’s evidence.

Illustrative only — not a result or a promise

What a teacher’s capstone evidence looks like

  • Class VIII Mathematics — Linear Equations in One Variable, one section of 35.
  • Before: 54% mastery on the chapter’s competencies; key weakness variable transposition; 17 of 35 needing remediation.
  • After a four-week practicum — diagnostic in week 1, named practice and one reteach in weeks 2–3, reassessed on the same competencies in week 4: 78% mastery; 6 of 35 still needing remediation.
  • The teacher decides what those six need next. The number is the teacher’s evidence for the teacher’s decision — not a product claim.

Parent communication

Written by the teacher, explained by the school

Parents hear about AI from the news before they hear it from the school. The session settles what the school says, who says it and when.

Teacher-written updates

A short note after a diagnostic or a reassessment, drafted with AI help from what the dashboard shows and read by the teacher before it goes. Nothing is sent to a parent automatically.

A 45-minute parent AI-awareness session

Run by the school at a PTM: what AI is, how the school uses it, what it never does, and how parents can support responsible use at home. Outline and materials are in the resources.

Plain words about the student AI tutor

The student’s AI tutor is part of the student’s AI plan; in the institutional programme the practicum classes are provisioned with it by the school or pilot. Every student, on any plan, gets homework, quizzes, doubts and progress. Say this plainly; parents ask.

Consent and questions

Parents know what data is held, why, and whom to ask. The policy names that person.

The 12-week implementation roadmap

Laid over your calendar in the room

The session ends with these twelve weeks dated against the school’s own exams, events and staff-development days.

  1. Week 0 · Leadership

    Principal & management orientation

    The 3-hour leadership workshop: what AI changes, what it must not replace, policy, data, the 12-week plan. Decide the cohort and the CPD recording approach.

  2. Week 1 · Teacher cohort

    Teacher AI foundations

    Modules 1 & 3 — understanding AI; the teacher’s role in an AI classroom.

  3. Week 2 · Teacher cohort

    Responsible AI + computational thinking

    Modules 2 & 4 — bias, privacy, integrity; CT across subjects.

  4. Week 3 · Teacher cohort

    Assessment & learning-gap workshop

    Modules 5 & 6 — decompose a real chapter; write competency questions; read an assessment for gaps, not marks.

  5. Week 4 · Teacher cohort

    Productivity lab + classroom workflow lab

    Modules 7 & 8 — hands-on with several AI tools; then the teach → assess → identify → diagnose → intervene → verify workflow on PrepGraph.

  6. Weeks 5–8 · Teachers with students

    Classroom practicum

    Module 9 — a 60-minute launch, then each teacher runs the workflow with one class, one chapter, 10–30 students. Weekly 30-minute clinic.

  7. Week 9 · Teacher cohort

    Intervention analysis

    Before/after evidence pulled from the dashboard; reteach decisions made; parent notes drafted.

  8. Week 10 · Teacher cohort + leadership

    Case-study presentations

    Module 10 — capstones presented to peers in the CBSE Format-F1 structure; feedback.

  9. Week 11 · PrepGraph Academy

    Assessment

    End-of-programme assessment; portfolio review; Level 3 decisions.

  10. Week 12 · Whole school

    Certification ceremony

    Certificates presented by the school; AI Champions nominated; CPD records finalised by the school.

Leaving the room

Decisions made, documents in hand

Nothing here is homework. Each decision is written with a name and a date before the session closes.

Decisions the leadership team makes

The cohort

15–40 teachers; which subjects and grades; which sections become practicum classes (one class, one chapter, 10–30 students each).

The CPD recording approach

Which sessions the school records under which CBSE/NPST domain, for how many hours, and who keeps the attendance and record sheet. The school decides; we supply the template.

The policy owner

One named person who owns the AI and responsible-use policy, answers parents’ data questions, and brings the draft back for review — first with the teacher cohort’s Module 2 “We will / We will not” clauses in Week 3, then on the review cycle the policy names.

AI Champion nominations

Two or three teachers who will mentor colleagues and carry the workflow forward — the future Level 4 AI Mentor Educators.

The calendar

Dates for Weeks 0–12 against the school’s exams, events and staff-development days; the PTM that hosts the parent session; the assembly or staff meeting that hosts the student and staff orientation.

What the principal leaves with

Draft AI & responsible-use policy

One page: purpose, permitted and prohibited uses for staff and students, data and privacy, academic integrity, owner and review cycle.

Policy template

Cohort plan

Names, subjects, practicum sections and the 12-week dates on the school calendar.

CPD record approach

Which session is recorded under which domain and for how many hours; who keeps the sheet.

CPD record sheet

Communication plan

What staff, students and parents are told, by whom and when — including the parent AI-awareness session and the student responsible-AI orientation.

Parent session outline

Independent programme, not a CBSE certification. PrepGraph Academy is not affiliated with, endorsed by, or accredited by the Central Board of Secondary Education. Certificates are issued by PrepGraph. The programme is designed around the CBSE 2026-27 training theme and the CBSE/NPST professional-development domains; the host school decides how to record the hours under CBSE’s CPD guidelines and keeps the attendance records — we supply the templates.

Start with the leadership session

Tell us your board, grades and teacher count. We’ll reply within two working days with a Week-0 date, the policy template and the cohort plan to fill in.