Your material
You own your book. You own what we make from it.
This is the question most people ask last and should ask first: what happens to the converted course and the question bank on the day you stop paying us. Here is the answer, including the part where the software does not yet have a button for it.
Ownership
Nothing you send us changes hands.
Your book was yours before you found this page and it stays yours. We take no ownership of it, no licence to publish it, and no right to use it anywhere except in the course we build for you. We do not need any of those things to do the work.
The converted course is yours as well. So is the question bank, including the questions our people typed in from your past papers and the page references underneath them. We wrote the software. We did not write your course, and we are not going to behave as though the conversion made it ours.
- Your original files
- Yours. We hold a copy so that every citation can be checked against the page it came from, and we hand it back or delete it whenever you say.
- The converted course
- Yours. The lessons, the modules, the text of every block, the label saying where that block came from, and the page reference under it.
- Your question bank
- Yours. Question text, the options, the correct answer, and the marker recording that a question was taken from your own papers word for word rather than written by anybody.
- Your students’ records
- Your school’s. Who enrolled, which lessons they opened, how they answered. We do not sell it, pool it, or look at it.
- Your name and colours
- Yours. There is nothing anywhere in what a student opens that names a supplier.
- The software
- Ours. That is the only line on this list that is.
We will not ask you to grant us a licence to your material as a condition of converting it, and there is no clause in what you sign that quietly does it anyway.
Copyright and rights
Before we convert a page, we ask who holds the rights.
This is the awkward conversation and we have it before any money moves, because having it afterwards is worse for both of us.
Plenty of good training material contains something the provider does not actually own. A sign chart lifted from the authority’s own booklet. A diagram from a textbook. A set of photographs bought for a printed handout under a licence that says printed handout. Putting that on a web page your students can open is a different act from putting it in a binder, and the difference is not one you want to discover from a letter.
- 01
Who wrote it
Did you write the material, or did somebody write it for you, and does the agreement say who owns the result? An instructor who left in 2014 and wrote four chapters is a real situation rather than a hypothetical one, and we would rather find it now than in year three. - 02
What is inside it that is not yours
We go through the book looking for anything that came from somewhere else and we ask you about each one rather than assuming. Official signage, statutory text, questions published by the examining body and stock photographs all have different answers, and three of the four are usually fine. - 03
What the licence actually permits
A licence covering a printed handout for students sitting in a room does not automatically cover a page on the open web. Where the answer is not clear, the safe move is to leave the item out and mark the gap for you, not to include it and hope nobody looks. - 04
If you do not hold the rights, we decline the work
Not a smaller quote. Not a disclaimer in the terms. Not a note in a footer nobody reads. We say so and we do not take the money. There is no version of this where we convert material you cannot show us the right to use.
What we will not do with it
Your material is not training data.
It is not going into a model, ours or anybody else’s, and it is not going to turn up inside a course we sell to the school two towns over. That undertaking covers material we hold today and anything you send us later.
An undertaking is worth something, and we would rather give you something better, which is what the code currently does. All three of these are checkable, and the first one is the reason the other two are easy for us to say.
- 1
No model is called at all
There is no language model anywhere in the conversion today. The interface for one exists in the codebase with nothing behind it, and no code path reaches a paid generation service. The conversion is an ingest pipeline plus people reading the result, which is slower and is most of what you would be paying for. - 2
Page reading is switched off
The service that reads scanned pages with no text layer is disabled in the deployment configuration, and while it is off the interface is not enabled on the project at all. Nothing about your book leaves the storage bucket it was uploaded to. - 3
We have no course of our own
Not one, in any subject. There is nothing for your material to improve, no catalogue it could be folded into, and no second customer who would benefit from it. The system is empty until somebody’s material is in it. That is the point of it.
When a model does become part of this, it will be named on the security page as a sub-processor before it reads a single page of yours, and we will still not use your material to improve it.
If you leave
What you get back, in what form, and how long you have to ask.
You can stop at any point, for any reason, and you do not have to tell us which one. What follows is what happens next.
local-academy-export/
│
├── originals/ the files you sent, unchanged
│ ├── theory-manual-2026.pdf 41.2 MB
│ ├── past-papers-2011-2025.docx 3.8 MB
│ └── scans/ 216 files
│
├── course/
│ └── course-version-3.json every lesson in order; each block
│ carries the label saying where it
│ came from and the page it cites
│
├── questions/
│ └── question-bank.json every question, its options, its
│ correct answer, and whether it was
│ taken from your papers word for word
│
└── MANIFEST.txt what is in here, who produced it,
on what date, and a checksum per fileThe originals are the easy half
The bucket your files sit in is versioned and has no rule that deletes a live object, and the part of the system that reads them cannot write to them. What you sent is what comes back, byte for byte, in the format you sent it.
The course is the half that needs explaining
What you get is structured files rather than a course that drops into another supplier’s app, because no interchange format carries a page citation and a per-paragraph origin label. Any developer can read them and nothing in them is obfuscated. What we are not going to do is call that a one-click migration, because it is not one.
| Cost of asking for your material back | Nothing |
|---|---|
| Notice we need from you | None |
| Reason we need from you | None |
| We confirm we have the request within | 2 working days |
| Files in your hands within | 10 working days |
| Kept after your account is closed | 90 days |
| Then | Deleted, and we write to tell you it is done |
| Who does the work | A person, by hand. There is no export screen |
Now the part most suppliers leave out
There is no export button. We went and read our own API before writing this paragraph, and not one of its endpoints exports anything: not a slow one, not a partial one, none. So on the day you ask, somebody runs the queries, collects the files, checks the result against the manifest and writes to you with a link to it. That is slower than a button and it is what exists. If the button ever ships it will appear on the updates page with a date against it.
Deletion has the same shape. There is no delete-my-school button either. It is a person running a script against the database and both storage buckets, which is exactly why we undertake to write and confirm it is finished rather than showing you a progress bar that would be a picture of nothing.
Everything in the table above is an undertaking we are making here, in writing, today. It is not a description of software behaviour, and we would rather tell you which is which than let you assume it is all automatic and find out in the worst week of the relationship.
We will not make getting your material back conditional on how the relationship ended, on a notice period, or on an unpaid invoice. If we ever want to argue with you about money, we will do it about money.
In one place
What we will not do with your material.
- We will not take ownership of your material, or ask for a licence to it beyond building your own course.
- We will not use your material to train a model, ours or anybody else’s.
- We will not use what we learn from your book to build a course we sell to another provider, including one down the road from you.
- We will not convert material you do not hold the rights to. We ask first, and we will say no.
- We will not sell your students anything, and there is nothing in what they open that names us.
- We will not hold your converted course back to win an argument about money.
- We will not tell you an export exists in the software while it does not.
Ownership settled, the next question is the practical one: what actually happens to the book between the day you send it and the day a course comes back. How it works
Send one chapter.
Not a demo of somebody else’s course. A chapter of yours, converted properly, so you can look at your own questions and your own page numbers and decide whether it is any good. Pick the chapter you know best. If it is no good, you have lost an email.
We will ask who owns the copyright before we convert anything. If you do not hold the rights to the material, we will say so rather than take the work.