The teachers' AI agent: lighten the preparation, give time back to teaching
Preparing materials, differentiating exercises, marking, writing reports and letters: these tasks absorb a considerable share of teachers' time — at the expense of what matters, the teaching relationship with pupils. Your AI agent absorbs that repetitive work. Hosted in France — on local inference or an isolated resource — pupils' data stays under control. The AI agent assists, the teacher decides.
Updated on
Differentiation: a scaffolded path (manipulation, guided diagrams) and an extension path (open problems). The exercises follow your schemes of work for the year.
For you to review and adjust to your class.
⛓ Source · your schemes of work + the official curriculum for the key stage
The assessment and the final wording stay yours — you correct, qualify and approve every comment.
✎ Action · drafts ready for review — the teacher approves
For teachers, a Blue Lemon Agent agent lightens the repetitive tasks — preparing materials, sequences and differentiated exercises, help with marking, letters and reports — to give time back to the teaching relationship. It runs on local inference or is hosted in France: pupils' data is never exposed to a foreign service, architecture designed to reduce exposure to extraterritorial legislation, location alone not being enough to guarantee immunity. The agent stays strictly in support: no automated decision about a pupil. Live within a few weeks. Teachers and the administrative team write to it from Microsoft Teams, Slack or their email, and families reach it on WhatsApp Business, the website chat or email — with no account to create and nothing to install. Reaching the school from the tool people already have means less non-take-up of rights and equal access to the service. These connections are included in every plan, at no extra cost, within the number of connections your level includes.
Reference points describing our offer, not results measured at a client. The scale of the gain is confirmed by a pilot on your own scope.
Why AI matters to schools — and why it worries them
Teachers want time for their pupils, but preparation, marking and administration take up a large part of it. Consumer AI promises relief — at the cost of pupils' data entrusted to foreign services, which is unacceptable for a public education service.
! The issue
The teacher is caught between the duty of continuity and equal treatment for every pupil, and a load of preparation, marking and administrative tasks that keeps growing. Yet most consumer AI tools amount to entrusting pupils' data, their work, observations and information about families to a third party, often hosted outside Europe and subject to the Cloud Act.
✓ Our answer
AI is only of interest to schools if it is sovereign and confidential by design. Local inference or an isolated resource hosted in France, systematic human oversight, decisions reserved to the teacher: the time saved on preparation is never paid for in lost confidentiality, nor in automated decisions about pupils. The aim is not to replace the teacher, but to give them back time for their pupils.
Protecting pupils' data: sovereignty & compliance
A class means personal data of minors and families, among the most sensitive there is. Here is how the architecture of our agents protects it, lesson after lesson.
Local inference
The agent can run on a resource at the school: no pupil's work or data leaves the network, nothing passes through a cloud.
Hosting in France
Otherwise, a dedicated and isolated resource, hosted in France under French law — pupils' data: processing and access within the European Union targeted by the architecture.
Reduced extraterritorial exposure
Exposure of pupils' data to the Cloud Act and FISA 702 is reduced by design; location alone does not guarantee immunity.
One isolated resource per establishment
No pooling of pupil data: an environment strictly dedicated to your school.
Encryption & controlled access
Encryption in transit and at rest, role-based access (RBAC), strong authentication and logging.
AI Act: governed deployment
The agent is strictly in support; no automated decision about a pupil; traceability and human oversight from end to end.
What depends on the architecture chosen These points are not general guarantees: they are settled deployment by deployment, in the quotation.
- The applicable location is that of the architecture set out in the quotation and verified before commissioning.
- Local execution is announced only for the configuration explicitly described and accepted in the quotation.
- The applicable isolation depends on the deployment mode set out in the quotation; no dedicated isolation is presumed.
- The encryption mechanisms in transit and at rest, their components and key management are those documented for the architecture chosen.
- Roles and permissions are configured and accepted for the identities and systems actually connected.
- The events logged, their content, their retention period and who may access them are defined for the deployment chosen.
See the agent at work
4 real situations, taken from those that come up most often. Pick one: the exchange unfolds as it would in your organisation.
A scripted demonstration. These exchanges show how the agent behaves — its sources, its refusals, what it leaves to your teams. Nothing is sent from this page, no model is queried here, and the matters named are fictional. That is precisely what we promise your data.
The behaviours shown here — monitoring, automation rules, routing and reminders — are configured with you during deployment, from your tools, your rules and your thresholds.
The architecture points named in these exchanges — location, local execution, isolation, encryption, role-based access, logging — are not a guarantee attached to the demonstration: they are those of the architecture set out in your quotation, and verified before commissioning.
The public body in this demonstration
Fictional public bodyCollège des Quatre-Vents — state lower-secondary school with 486 pupils (fictional school)
- Sector
- State lower-secondary school covering the four years to age 15 — 18 forms, 11 subject teams, a reception unit for newly arrived pupils with little French, 36 teaching weeks a year
- Headcount
- 41 teachers — 18 of them form tutors —, 4 pastoral staff and a two-person office; the 41 teachers are the users of the agent
- Public served
- 486 pupils and 431 families; 12 pupils with little French received this year; 5 languages other than French declared at enrolment over the last three years
- Order of magnitude
- 738 teaching sequences prepared a year — 18 per teacher; 540 sets of report comments a year (18 forms × 10 subjects × 3 terms), that is 14,580 individual comments; 54 form review boards; 9,400 messages from families received in the year
- Tools in place
- The school's digital workspace, the pupil-management software (lesson log, marks, absences), the 11 team schemes of work, the office's letter templates and Microsoft Teams for internal exchanges — the agent plugs in read-only, nothing is replaced and nothing is migrated
- Who decides
- Each teacher validates, corrects or discards anything that carries their name — no comment goes out without them; the form tutor settles the report text; the review board remains sovereign; the head teacher settles what is published to families and signs the school's letters
- Points for improvement
- A differentiated three-lesson sequence takes 5 hours, and only 246 of the year's 738 sequences actually carry both pathways; 1 h 15 per form per term on writing report comments alone, that is 540 sets in the year; 6,400 of the 9,400 messages from families are about 11 recurring questions; and 62 of the 431 families never open the digital workspace
At Collège des Quatre-Vents, 41 teachers prepare 738 sequences a year, write 14,580 report comments and receive 9,400 messages from families, with no extra staff. It is not the teaching that eats the time, it is everything around it: writing the same exercise twice at two levels, hunting for the homework where an observation was noted, answering the question about pick-up times for the eleventh time. The exchanges below span a school year, from reading the lesson log to the report presented to the governing board. The agent runs on local inference on a machine in the school, reads the lesson log, the schemes of work and the marks entered without ever writing to them, and no output concerning a pupil exists before the teacher has signed it.
This public body, its figures and the exchanges that follow were invented for the demonstration. They illustrate a common situation; they describe no real service.
What the count says:
· 246 of the 738 sequences actually carry both pathways — one in three. The other 492 were prepared with no written support or extension route, and that is not a teaching choice: a differentiated three-lesson sequence takes 5 hours, and those five hours do not exist three times a week.
· 6,400 of the 9,400 messages from families are about 11 recurring questions — timings and pick-up, canteen, absences and notes, equipment, school trips, appointments with a teacher, access to the digital workspace, grants, transport, the year-10 work placement, the report and board calendar. Two messages in three already have their answer written somewhere in your school rules or your staff notices.
· 3 of your 11 team schemes of work still carry last year's ordering, although the teams changed it at the start-of-year meeting on 2 September — the minutes are in your digital workspace. Those 3 schemes govern 214 sequences of the year.
What I bring you this morning, already written: the 3 schemes are rewritten in their 2 September version, with the replaced line and the minute line that replaces it set side by side; the 11 answers to families are drafted, sourced on your school rules and staff notices, each with its date; and a complete three-lesson sequence on fractions for the first year, with its support pathway and its extension pathway, is waiting for you.
The gain, costed on your own volumes: preparing a differentiated sequence goes from 100 % to 18 % of the time it takes you today — from 5 hours to 54 minutes, and those 54 minutes are reading and adapting to your class, not writing.
The next step, and it takes fifty-four minutes: you read the fractions sequence, you correct what does not suit your class, and it is used on Monday morning. preparing-the-year_738-sequences-11-questions.pdf246 differentiated sequences out of 738, 6,400 messages on 11 questions
⛓ Sourced · lesson log for the last 3 years, 11 team schemes of work, minutes of the start-of-year meeting of 2 September, 9,400 messages from families
What the count of your lesson log says:
· Mathematics: 168 sequences a year, 7 teachers — and it is one of the 3 schemes that still carried the old ordering.
· French: 154 sequences, 6 teachers.
· History and geography: 132 sequences, 5 teachers.
Those three teams account for 454 of the 738 sequences: three teams out of eleven, 61 % of the school's preparation. The other eight open in January, two per fortnight, with nothing left waiting.
And I would advise against opening all eleven at once, with the figures behind it: the first three teams need three thirty-minute meetings — one per team, to settle the reference scheme of work. All eleven need eleven, that is five and a half hours to fit into timetables that do not overlap — three weeks' delay to going live for 39 % more sequences. Three teams in September, eight in January: the first sequence is used on Monday, and the whole year is covered without a single meeting being added to your calendar.
What those three meetings cost you, and what they give back: thirty minutes per team to settle the reference scheme of work and to tell me what you allow me to reuse from the last three years of the lesson log. Walking out, each teacher has in front of them the list of their 18 sequences for the year, each with the number of hours it took them last year and what it will take this time — 5 hours against 54 minutes.
What that will give over the year, and it is the figure I will defend in front of you in June: 612 of the 738 sequences prepared with me — 454 for the three September teams, 158 of the 284 sequences of the eight teams opened in January.
The next step I propose: that the three thirty-minute meetings be put in the diary this week, and that I compare your 11 schemes of work with the minutes of the latest team meeting every night. Your schemes will stop ageing in silence, and the September read-through will not have to be done again: it becomes a five-minute validation at each team meeting. preparing-the-year_738-sequences-11-questions.pdf454 sequences for 3 teams, 612 over the year, opening order
⛓ Sourced · count of sequences by subject team over 12 months, 11 schemes of work, timetables of the 41 teachers
Local inference means the model computes on your machine: the text of a piece of work, an observation recorded about a pupil or a family's message never crosses an outside network to be processed. If the school would rather not administer a machine, the other route is an isolated resource hosted in France, dedicated to Collège des Quatre-Vents, shared with no other school.
What that changes, point by point:
· Pupil data trains no model, neither ours nor a third party's.
· I work read-only, and the technical account I read through has no right to write into your pupil-management software — that can be checked with one command, which is worth more than a promise.
· Encryption in transit and at rest, role-based access — rights follow the role: a teacher opens the forms that are theirs, and nothing else. 41 roles for your 41 teachers, and the log shows 0 out-of-role access since go-live.
· Hosting in France, under French law, architecture designed to reduce exposure to extraterritorial legislation, including against an American operator hosting in Europe — location alone not being enough to guarantee immunity, the exposure is documented for the configuration chosen.
· A complete log: who asked what, when, and what the system produced.
And on decisions, this is not caution, it is law: a decision producing legal effects concerning a person cannot be based solely on automated processing — that is Article 22 of the General Data Protection Regulation. A mark, a progression to the next year, a guidance decision are decisions; they remain those of the teacher and of the review board, and that is exactly what makes them reasoned and contestable in front of a family. The route that gives you the result without giving anything up is the one we are taking: I prepare, I cross-check, I set the entries alongside, and you sign in fifteen seconds instead of writing for fifteen minutes.
What I can do beyond that, the day the school decides to: having an AI system assess learning outcomes falls under the education heading of Annex III of the European artificial intelligence regulation — high risk, and not prohibited. So it is a compliance programme, not a wall: technical documentation, logging, named human oversight, information of the persons concerned. It is written and I put it on the table the day you want to go there; as long as the agent stays in a supporting role, you do not need it.
And the pupil, like the family, knows who they are talking to: I introduce myself as a digital assistant of the school in the first sentence, because the European artificial intelligence regulation requires that anyone interacting with an AI system be informed — and a teacher can be called in at any moment, in which case I pass the request on with the subject already noted.
The figure that sums all of this up: 0 pupil data out of the school network over the year, and 0 transfer outside the European Union.
What I propose: that I keep up to date the record the academy's data protection officer will ask you for — hosting, data processed, retention periods, who accesses what, and the mention of local inference. The first version is written and attached; it updates itself each time a new use is opened. technical-framework_where-pupil-data-lives.pdfLocal inference, read-only, 0 transfer outside the EU
✎ Framework · deployment architecture, technical account rights, first version of the processing record, Article 22 GDPR
What you receive, lesson by lesson:
· Lesson 1 — the fraction as a share. Objective written in one sentence, timed run-through over 55 minutes, materials listed, written summary drafted, and the two prompts to have ready when the class gets stuck.
· Lesson 2 — locating and comparing. Same shape, with the mistake your pupils have made most often on this topic over the last three years, taken from your lesson log, and the exercise that tackles it head-on.
· Lesson 3 — solving a problem. Run-through, group management, and the 15-minute formative assessment with its mark scheme.
The two pathways, both written out in full:
· Support — 18 exercises: fraction strips to handle, diagrams already drawn to be completed, instructions in one sentence and one only, numbers chosen so the arithmetic does not hide the concept, step-by-step corrections.
· Extension — 16 exercises: open problems with several ways in, written justification required, two situations where the intuitive answer is wrong, and an extension towards proportion.
Where it comes from, and you can check it line by line: your mathematics team's scheme of work in its 2 September version — the one I brought up to date —, your lesson log for the last three years, and the documents you yourself put in the digital workspace. Every exercise carries the line of the scheme of work it answers: you can see at a glance whether the lesson sits in the right place in the year.
The gain, costed: preparation goes from 100 % to 18 % of the time it takes you — from 5 hours to 54 minutes. Over the 612 sequences we will prepare together this year, that is 2,509 teacher hours given back to the classroom.
And the figure that does not flatter me, because it is better that it comes from me: on the first batch of 96 sequences in September and October, 22 had to be rewritten on substance — 23 %. The cause is measured, not assumed: 17 of the 22 rested on a topic the team scheme of work had moved to the following term, and I was faithfully reproducing a scheme nobody had updated. What I did about it: I now compare the 11 schemes of work with the minutes of the latest team meeting every night, and any scheme a meeting has just changed is rewritten overnight. The next batch: 4 sequences rewritten out of 108 — 3.7 %, and all 4 concerned forms with a particular profile, not a single out-of-date scheme left.
The next step I propose: that I prepare the next 6 sequences of your scheme of work in advance and that you read them in pairs, in thirty minutes, once a fortnight. You stop preparing next week's lessons on Sunday evening: you are a term ahead, and you did not pay for it in hours. fractions-sequence-year-1_three-lessons-two-pathways.pdf3 lessons, 34 marked exercises, 5 h → 54 min, 23 % → 3.7 %
⛓ Sourced · mathematics team scheme of work of 2 September, lesson log for the last 3 years, log of sequences rewritten over two batches
What is ready, and costs you no extra hour:
· Each of the 34 instructions exists in an easy-read version — short sentences, one message per sentence, everyday words rather than school words. Both versions carry the same concept and the same exercise: it is not an easier exercise, it is the same step, with a handrail. You hand out one, the other, or both side by side.
· No easy-read version goes out without the sign-off of the teacher who uses it, and that sign-off costs you no preparation: the two versions arrive side by side, the concept assessed highlighted in each — forty seconds per instruction. Of the 34, 29 were signed off as they stood and 5 came back to me: that is the figure that does not flatter me, and in all five cases my short sentence had dropped a constraint of the question — « in one sentence », « without a calculator ». What I did with that: the constraints of the question are now copied word for word before the rewrite, and across the 214 instructions of the term, 8 came back to me, 3 of them for that reason — from 5 out of 34 to 3 out of 214.
· The adjustments recorded in a pupil's file are decisions of the school, not mine: I reproduce them to the letter — extra time, font and line spacing, question read aloud, shortened paper —, and I build the compliant document in thirty seconds instead of a quarter of an hour.
And if you want me to go further, I do it as soon as you ask: I can propose further adaptations, written out, each with the concept it targets and the exercise where it applies. They are proposals put on your table; putting them in place is a teaching decision, and it belongs to you — which is what makes it defensible in a support meeting.
What the term produced, measured on what was handed in, never on what a pupil is worth: 214 instructions handed out in easy-read form, 206 signed off by the teacher without a single return, and 7 of the 12 pupils with little French handed in all the work set, against 4 out of 12 the previous term — the short version is in plain French, and that is the one they handed in. That is a count of work handed in, not a measure of attainment — and it is the only figure I need to say that the instruction was the wall.
The next step I propose: that the easy-read version becomes the default format for the instructions of every assessment in the school, with the long version kept attached. It is written for the 34 exercises of this sequence; extending it needs a decision of the teaching team, and that takes ten minutes at the teaching council — the paper for the meeting is already drafted. accessibility_easy-read-214-instructions-signed-off.pdf34 in easy-read, 5 back then 3 out of 214, 206 signed off with no return
⛓ Sourced · sign-off log of the 214 easy-read versions, adjustments recorded in pupil files, log of work handed in over the term
Where those texts come from, and there is nothing else in them: the marks you entered in the pupil-management software, the observations you recorded in the lesson log, the remarks you wrote on the work handed in. Every draft carries, in the right-hand column, the entries it derives from — the date, the mark, the observation, word for word. You check in fifteen seconds instead of searching for a quarter of an hour.
The figures on the report are yours, reproduced exactly and checkable line by line. What I make is the text — and it says nothing your entries do not already say. A comment talking about something you never wrote down would be indefensible in front of a family; this one defends itself entry by entry.
The gain, costed on your volumes: writing a form's report comments goes from 50 % to 12 % of report time — from 1 h 15 to 18 minutes per form per term. Over 540 sets of comments in the year, that is 513 teacher hours given back. The remaining 1 h 15, where you reread your marks and make up your mind, does not move by a minute: that is the job itself.
The honest figure, and it is the one I want you to look at: in the first term you corrected 1,293 of the 4,860 drafts — 27 %. That is exactly what I expect. A comment a teacher has not touched is not a comment, it is a template — and what I give you back is the 3,567 you validated with a word, plus the 1,293 where you started from a text instead of a blank page.
On measuring people, I give you the mechanism rather than a principle: the day « comments reworked per teacher » becomes a tracked figure, it becomes a target — people rework for the sake of reworking, the sentence changes without the judgement changing, and the indicator stops measuring the quality of the report and starts measuring the reaction to the indicator. So I offer it by form and by term, never by teacher. One exception, and it is not one: whoever signs is named. The report carries the teacher's name and the board's opinion carries the form tutor's — a signature is not a counter, it is what makes the text binding.
The next step I propose: that the drafts be put up five days before entry rather than the day before. In the first term, teachers who had them five days ahead validated in 14 minutes; those who had them the day before, in 26. The calendar is written and it fits inside yours without moving anything. report-comments_28-drafts-with-their-entries-alongside.pdf1 h 15 → 18 min, 3,567 validated, 1,293 corrected — 27 %
⛓ Sourced · marks and observations entered by teachers, lesson log, validation log for the first term
Before the sitting, one page per form, ready two days ahead: for each pupil, the marks entered, the validated comments, the absences recorded, the adjustments on file and the points left open at the previous board. The form tutor arrives with the material assembled — and of the 54 boards this term, 51 started at the scheduled time, against 34 last year.
After the sitting, the minutes built on the school's template: what was said in the sitting, and nothing else. An opinion that was not expressed does not appear in them; the draft reaches you with the pending passages marked in plain sight, and you complete them or remove them. 45 minutes of writing down to 8 minutes of reading, and the minutes go out the next morning instead of three weeks later. I do not count these hours in the report I will present to you in June: they are not among the three items we measured, and an unmeasured gain has no business in a total.
Parents' evenings, right through to the end of the act: I offer the slots you have opened, and only those — never a slot you have not declared —, I book the appointment, I send the confirmation in the standard version and in the easy-read version, the list of documents at the top, and I remind them the day before. Over the year's two evenings: 1,240 appointments booked, 92 % kept against 78 % last year, and 61 families who had never come, came.
And the deadlines that always fall at the wrong moment: year-10 work placements, trips and consent forms, textbook orders, marks returns, disciplinary boards to convene in due form. Every deadline carries its date and its recipient, the alert falls ten days ahead, and the matching letter is already drafted on the office's template. Over the term, 0 deadline missed out of 63 tracked.
The next step I propose: that the preparation page be sent to the 18 form tutors 48 hours before each board, automatically. It is the only thing I would send unprompted on this perimeter, it goes only to the teachers concerned, and the head teacher withdraws it with a word. review-board_preparation-minutes-and-appointments.pdf45 min → 8 min, 1,240 appointments, 0 deadline missed out of 63
⛓ Sourced · calendar of the 54 boards, the school's minutes template, slots declared by teachers, deadline register
Where I answer, with no family having to create an account or install an app: WhatsApp Business, the chat on the school website and e-mail. And on the staffroom side, you write to me from Microsoft Teams, Slack or your mailbox, without changing tools; oversight runs from a web dashboard. These connections rest on open standards, including the MCP protocol; they are included in every plan at no extra cost, within the number of connections your tier includes. Only the charges billed by the platforms themselves — WhatsApp Business bills per conversation — are passed on at cost, with no margin, outside the subscription.
The rule that protects the school, and it is simple to hold: the exchanges are about organisation and procedures — timings, trips, canteen, absences, appointments, grants, the year-10 work placement. Anything touching on the follow-up of a pupil goes back to the teacher and to the school's own channels, with the appointment already offered and the subject already noted — the family does not have to explain twice, and the teacher receives a request that is ready.
The gain, costed: answering a recurring parent question goes from 17 % to 8 % of the time the request takes — from 3 minutes 24 to 1 minute 36. Over 6,400 messages a year, that is 192 hours given back to teachers, and a response time that goes from 4 days to a few minutes, Saturdays included.
The figure that has nothing to do with hours, and that may decide your governing board: 62 of your 431 families never opened the digital workspace. 41 of them asked their first question on WhatsApp this term — that is non-take-up avoided, and equality of access that can be counted.
And you keep the key: I offer you a written mandate, capped to the 11 questions listed by name, dated, reviewed after one term and withdrawable with a word from the head teacher. Outside those 11 questions, the answer reaches you drafted and sourced, and you are the one who sends it. The text of the mandate and the information notice to families — the one to be published on the website and displayed at reception — are both drafted.
The next step I propose: that the school's WhatsApp number go out in the home-school planner at the January return, with the 11 questions on the cover and a statement that a digital assistant answers first. The insert is written, it fits on half a page, and it costs one signature. families_eleven-questions-on-every-channel.pdf6,400 messages, 3 min 24 → 1 min 36, 41 families reached
✎ Framework · mandate drafted, list of the 11 questions, information notice to families, channel log over one term
The calculation, item by item, so that anyone can redo it:
· Preparing a differentiated sequence: 612 sequences, 5 hours down to 54 minutes — 100 % → 18 % of the task's time — that is 2,509 hours.
· Writing a form's report comments: 540 sets, 1 h 15 down to 18 minutes — 50 % → 12 % of report time — that is 513 hours.
· Answering a recurring parent question: 6,400 messages, 3 minutes 24 down to 1 minute 36 — 17 % → 8 % — that is 192 hours.
What those 3,214 hours represent on the statutory working-time basis — 1,607 hours a year, 151.67 hours a month: exactly two working years, spread over 41 teachers: 78 hours each over the year, a little over two hours per teaching week. No post cut, no post created: it is teacher time given back to the classroom, at identical staffing.
What I did not put in the total, and I would rather tell you: the 54 sets of board minutes down from 45 to 8 minutes, the 63 administrative deadlines tracked, the preparation of the boards. Those are real gains, but they are not among the three items we measured — a total inflated with what was not measured cannot be defended in front of a governing board.
What those hours became, according to your own records:
· Sequences carrying both pathways: 246 out of 738 → 612 out of 738.
· Time to answer a family: 4 days → a few minutes, Saturdays included.
· Board minutes sent: three weeks → the next morning.
· Parents' evening appointments kept: 78 % → 92 %, and 61 families who had never come, came.
· 41 of the 62 families who never opened the digital workspace asked their first question.
· 7 of the 12 pupils with little French handed in all the work set, against 4 out of 12 the previous term.
The figure that does not flatter me, published with the rest: 22 sequences rewritten on substance out of the 96 in the first batch — 23 %, down to 4 out of 108 — 3.7 % once the schemes of work were brought up to date every night. And 1,293 comments corrected out of 4,860 — 27 % —, which is the proof that the text remains the teacher's.
And the framework measures: 0 decision about a pupil taken without a teacher, 0 pupil data out of the school network, 0 transfer outside the European Union, over 19,300 logged outputs.
What I propose for the meeting: the calculation page is written and fits on one side — three lines of arithmetic, six response times, three framework measures. Attach it to the notice of meeting: a figure read the day before is discussed better than a figure discovered in the room. year-report_3214-hours-given-back-to-teaching.pdf100→18, 50→12, 17→8, and the arithmetic redoable on one side
⛓ Sourced · lesson log, sequence and comment logs, channel records, deadline register
· I compare your 11 schemes of work with the minutes of the latest team meeting every night, and I write the up-to-date version of any scheme a meeting has just changed. It is that action that took substantive rewrites from 23 % to 3.7 %, and it costs five minutes of validation per meeting. And the reverse is true too: a scheme the team withdraws leaves my sequences at the same hour — I never serve a document the team has decided to withdraw.
· I send the 18 form tutors their preparation page 48 hours before each review board. It goes only to the teachers concerned, it does not leave the school, and the head teacher withdraws it with a word.
· I flag a subject that is rising before it fills your mailboxes. This term, questions about the year-10 work placement tripled in a fortnight, to 340 messages — the half-page notice for the home-school planner and the message to the 118 year-10 families are written and waiting for you.
On measuring teachers, the answer is clear, and it is not a moral one: it is law first, then mechanics.
The law first: measuring staff activity is not prohibited — an employer has the power to frame and monitor the activity of its staff, under three cumulative conditions: proportionality, prior information of the persons concerned, and information of the staff representative bodies before implementation. So if the head teacher asks me for an activity indicator, I produce it — and I bring the three conditions already met: the purpose written down, the staff information notice drafted, the retention period fixed. The decision is theirs, and I hand it to them in ten minutes instead of a file.
The mechanics next, and they are what should decide: the day « comments reworked per teacher » or « sequences prepared per teacher » becomes a tracked figure, it becomes a target — people prepare for the counter, they rework for the counter, and the instrument stops measuring teaching and starts measuring the reaction to the counter. So I offer the indicator by form, by subject and by term: it shows what the school produces, it can be presented to the governing board without naming anyone, and it actually decides something — where to put the effort next term.
One exception, and it is not one: whoever signs is named. The report carries the teacher's name, the minutes the form tutor's, the letter to families the head teacher's. A signature is not a counter: it is what makes an act binding.
The record by form is ready, attached, and it updates itself every Monday. what-the-agent-does-unprompted_and-what-is-logged.pdf3 reversible actions, indicators by form and by subject
✎ Framework · configuration of the unprompted actions, send log, lawfulness conditions for an activity indicator
What there is to dismantle the day you stop:
· The index. It is deleted, and it contains no pupil file — only what is needed to find your documents where they already are. Your pupil-management software, your digital workspace and your lesson log have not moved by a single byte.
· The log of requests and outputs. It is handed to you in an open format, or destroyed — the school chooses, and the question is settled at go-live, not on the way out.
· The 612 sequences, their 34 exercises each, their easy-read versions, the 11 answers to families and the 11 schemes of work brought up to date. They are at the school: they are made of its own material, they stay in its files, readable without us, in the office formats you already use. That is the teaching capital this year will have created, and it would be improper for it to stay with us.
What does not exist, and what should be checked with every supplier: no migration on the way in, therefore no migration on the way out. Your tools stay yours, no format belongs to us, and your 41 teachers work with the same software as before.
On retention, since the data protection officer will ask: outputs concerning a pupil are kept for the current school year and the following one, then erased; the period is written in the processing record and changes with one line.
What I propose so that this does not stay a sentence: a dry-run exit at the end of the first term, half a day: we switch off, we check that preparation, reports and letters work exactly as before, we switch back on. The protocol is written, it fits on one page, and the date that costs you least is the Wednesday afternoon of the second week of February — 4 sequences and 0 review board in that slot, against 38 sequences and 6 boards on a Monday at the end of term. The governing board will know what the promise is worth before committing to a second year. technical-framework_where-pupil-data-lives.pdfReversibility: 0 migration in, 0 migration out
✎ Framework · index architecture, export formats for the outputs, retention periods, dry-run exit protocol
Your case is not here? That is exactly what a 15-minute conversation is for. Book the free audit →
The uses of AI in support of teachers
Each use corresponds to an agent we deploy. All work in support, subject to your approval, with no automated decision about pupils.
Preparing materials & sequences
Generation of sequences, worksheets and lesson materials aligned with the curriculum and your schemes of work — for review and adaptation.
Differentiated exercises
Production of exercises at several levels (scaffolding and extension) to answer the range within the class.
Help with marking & feedback
Draft feedback and comments from your observations; the teacher assesses, qualifies and approves.
Letters & reports
Drafts letters to families, class council meeting minutes and notes using your templates.
Plain language & easy-read draft
Preparation of a plain-language version and a draft easy-read (FALC) transcription, subject to human validation.
Answers to families
Answering parents' recurring questions: organisation, procedures, practical information about the school.
Organisation & diary
Preparing class meetings, scheduling meetings with parents and tracking administrative deadlines.
On quote View the agent page →In 15 minutes we identify the agent that will give your staff the most time back — without oversizing the project.
How much time can a teacher win back?
By lightening the preparation of materials and the production of feedback, a teacher can aim for a noticeable reduction in the time spent on repetitive tasks — reinvested in the teaching relationship and in supporting pupils.
The stages of your AI agent project
Audit & scoping
15 minutes to target the use case with the best return.
Quote or direct sign-up
A catalogue offer is bought online; a specific need gets a costed quote.
Design
We design the agent and its guardrails.
Integration & testing
We connect your tools to the agent, which is itself hosted in France.
Rollout
Going live and training your team.
Operation
Continuous supervision and improvement.
An agent in support of teaching teams
A teaching agent (preparation, help with marking, administration), installed and operated for you. Choose according to how the school is organised. Prices exclude VAT — annual subscription, the time it takes for the gains to settle in.
Setup + controlled subscription
- Installation, configuration and training for your teams
- Operation, human oversight, updates and support
- Sovereign hosting in France, a dedicated and isolated resource
All inclusive, no setup fee
- Setup included (installation, configuration, training)
- Operation, human oversight, updates and support
- Sovereign hosting in France, managed end to end
On site, you own it
- Hardware installed on your premises (you own it)
- French / European AI models run locally
- Secure remote maintenance (Pro support included)
Four guarantees that matter to a school
Your questions, our answers
Does the agent assess the pupils?
Is pupils' data protected?
Does it fit the framework of the state education service?
Can AI really help prepare lessons?
Do we have to change tools or digital environment?
How long does it take to deploy an agent?
Which tools can families use to reach the agent?
Other public service professions in support
Let's size up the potential in your school
15 minutes to identify the use most helpful to teachers — hosted in France, supervised, with no commitment.