How to pilot AI in a school: start with one class

AI School builds on the school’s existing courses, classes, teachers and timetable, and does not ask anyone to change how they teach first. It usually starts with one pilot class or one pilot course and moves through four stages; each stage spells out what is delivered, what the school needs to provide and how sign-off works, and the school expands only after a stage passes.

Talk to us about your school Back to the AI School overview

The short version

No need to change teaching first: it sits on top of the school’s existing courses, classes and timetable.
Start small: a pilot class or course goes first, then more subjects, classes and campuses once it has proven useful.
Every stage can be signed off: deliverables, what the school provides and the sign-off criteria are written down before the stage begins.
Accounts are managed by the school: AI School is set up separately for each school and kept apart from the personal OneOneTalk app.

🌐 About language: This page introduces AI School in English. AI School is currently delivered in Chinese; if your school works mainly in English or another language, say so in your enquiry and we will tell you plainly what is and is not available.

The four stages

Stage	What is delivered	Who usually joins first
Stage 1 · Setup	School admin console, accounts, role permissions, classes and course materials; existing rosters imported or synced; basic reports switched on	Academic office and the staff member who looks after school IT
Stage 2 · Teacher productivity	Teachers use AI to plan lessons, create practice, mark homework and run classroom activities; confirmed content and classroom observations feed into learning reports	Subject teachers in the pilot class
Stage 3 · Student learning	Student access opens: course questions, revision tasks and learning summaries; a first set of Student 360 profiles is reviewed by teachers, who draw up support plans	Students and the form teacher in the pilot class
Stage 4 · School review	Leadership dashboard, class pages, class profiles and progress reports; progress reports can be turned into a version written for parents	Principal, academic leads and year-group leads

What to prepare before stage one

A pilot scope: one class or one course, ideally with one or two teachers who want to try it.
Course materials: learning objectives, textbooks and any existing slides and practice for the pilot course.
Rosters and roles: teacher, student and class lists for the pilot, and who is the form teacher and who teaches each subject.
A school-side lead: someone to coordinate teachers and sign off each stage.
Sign-off criteria: agree in advance what counts as done — for example, whether teachers used it in real lessons, and whether learning reports were used to adjust teaching.

Four places where school leaders most often get stuck

Launching everywhere at once: when the whole school goes live together, teachers have no time to adapt and nobody can tell where problems come from. Get one class working first, then expand.
Buying a tool without settling the workflow: if AI-generated content skips teacher review, it is soon abandoned or misused. AI School makes teacher confirmation a required step before anything is used.
Sorting out data access later: who can see which students’ data should be decided in stage one, together with role permissions, not patched after something goes wrong.
Skipping the review: without a review, there is no basis for deciding whether to expand. The stage four dashboard and progress reports exist for exactly that decision.

From first email to stage one

Email us with the size of your school and the year group or subject you would like to start with, and we will agree the scope, materials and sign-off criteria for the first stage together. Data imports are validated with clear error prompts; once existing rosters are imported or synced, classes and roles can be set up.

Which modules are switched on is agreed with each school as part of its delivery scope. After each stage passes sign-off, the school decides whether to extend to more subjects, classes and campuses.

This page describes what AI School does and how we work with schools; which modules are switched on is agreed with each school as part of its delivery scope.

Questions schools ask

Does the whole school need to start at once?

No, and we advise against it. Schools usually start with one pilot class or course and extend to more subjects, classes and campuses after each stage is signed off.

How is each stage signed off?

Before a stage begins, the deliverables, what the school needs to provide and the sign-off criteria are written down and agreed by the school and us; the next stage starts only after this one passes.

Can we import our existing student rosters?

Yes. Existing rosters can be imported or synced, and imports are validated with clear error prompts. The exact integration depends on the school’s configuration and delivery scope.

Who do we contact first, and what should we prepare?

Email ceo@oneonetalk.com with a short description of your school and the year group or course you would like to try first. We will agree the scope, materials and sign-off criteria for the first stage together.

Want to try it in your school?

Tell us the size of your school and which year group or subject you would like to start with. We begin with one class and agree the scope, materials and sign-off criteria for the first stage together.

Enquire about AI School See the staged rollout