Privacy Policy

What we collect, why, and what you can ask us to do about it

Mitsuru Technologies Ltd

Last updated 15 October 2026.Print or save as PDF

1. Who Is Responsible for Your Data

Mitsuru Technologies Ltd is the data controller for the personal data described in this policy, except for student work on the learning platform, which is dealt with in section 3. We are registered in England & Wales under company number 16755264, at 11 Poppy Mills Road, Brockworth, Gloucester, GL3 5AB, United Kingdom.

For anything about your data, including a request to see it or to have it deleted, email [email protected]. We have not appointed a data protection officer, because we are not required to: your request goes to the person who runs the company.

2. What We Collect, and Why

Every purpose below has a lawful basis, which is what the UK GDPR requires us to tell you. Where the basis is consent you can withdraw it at any time, and where it is our legitimate interests you can object, see section 10.

We do not buy personal data, and we do not sell it. We ask for a work email address for an early access enquiry because that is what tells us an organisation is asking, not to build a profile of you.

Special category data

We do not ask for special category data (health, disability, beliefs, and the rest of the Article 9 list), and we do not record it. If something of that kind comes up in a coaching conversation, it stays confidential under clause 9 of our Terms & Conditions and does not go into our notes.

There is one exception, and it only happens if you ask for it: where you want us to take something into account, an adjustment for a disability, say, we will record only what is needed to do that, and only with your explicit consent, which is the Article 9(2)(a) condition we rely on. We will ask for it in writing, tell you what will be written down, and delete it if you withdraw consent, which you may do at any time without giving a reason.

3. Students Using the Learning Platform

Where a college uses the platform for its students, the college is the data controller and we are its processor. The college decides who has an account, what work is set, which of its staff may see it, and how long it is kept. We act on its instructions under our agreement with it, and we do not use student work for our own purposes, to train anything, or to market anything.

If you are a student and want to see, correct, or delete your data, you can ask us and we will pass the request to your college, but it is the college that decides and answers. Your college’s own privacy notice covers what it does with your data.

We remain the controller of the records we need in order to run the platform at all: the address a sign-in code goes to, the passkeys registered to an account, and the security logs. Staff access to student work is logged, and the student sees the same log the staff do.

We email students about their own coursework. When a task is set for you, when your work is marked or sent back, when a teacher comments on it, or when you are put on a module, we send a message to the address your college holds for you saying so and linking to it. It is the same thing the site shows you when you sign in, sent so that you find out without having to. It is not marketing, we do not send anything else, and you can turn it off at any time from your modules page without losing anything: what is waiting is still shown on the site either way.

Environment variables you give a deployed app, such as an API key for a service your project uses, are stored encrypted, are shown to nobody including your teacher, who sees only their names, and are used for one thing: starting your app. A teacher marking your work can start your app with them without ever being shown them. Deleting the saved deployment deletes them, and they go with your account when it goes.

Deploying without an account makes a guest account in your browser: an account with no name and no email address, holding what you deployed and any database you made, kept for a day after you last use it or until your browser’s session has ended, whichever is first, and then deleted with everything in it. Nothing is asked of you to make one but that you pass the same spam check as the forms on this site. Giving it an email address makes it an ordinary account, described above.

A database you make for a deployed app is stored as its name, its kind and a generated password, encrypted the same way, and as whatever you put in it, saved as a file when it stops. Only an app you deploy here can reach it. A teacher marking your work can start it with your data and cannot change what is saved. Deleting the database deletes the data, and both go with your account when it goes.

Git repositories, for staff and students of a college that uses the platform, are kept on a Git server we run on our own machine in the United Kingdom, and copied every night to our storage in the AWS London region: what you push, its history, and your name as the platform knows it, with any large files and attachments beside them. Every repository is private, and only the people its owner adds can see it. Your email address is not sent to it. Removing or suspending an account switches its Git off straight away, and its repositories are kept until the college, or you, ask us to delete them.

Coursework in Git. Where a teacher lets students clone a task, each student gets a private repository for it on the same Git server, which they and the module’s teachers can see. What a student pushes there is their coursework, kept as described above, and a copy of it is also kept with the rest of their work on the platform so it can be run, marked and handed in. Turning cloning off stops their access to the repository; it deletes nothing.

A free lesson from our Resources pages is not this. Nobody needs an account to read one, and if you make one in order to send answers, there is no college involved at any point: we are the controller, and the request routes above do not apply because there is nobody to pass them to. Write to us and we answer it ourselves. Section 13 says what those accounts hold.

Teachers can compare a class’s saved code, and look at how a student’s code was built. On a coding task, a member of staff who teaches the module can ask the platform to compare each student’s saved copy with the others’ by the shape of the program, and to read the versions a student’s copy was saved in (every autosave is kept) to see whether it grew the way something being worked out grows. Both are evidence for the teacher’s own judgement, in the words “similar to” and “looks like” and never a finding about you:nothing from either is shown to a student or written to a student’s record, and that a teacher ran or opened one is recorded in our administration log. Where the college has turned AI on and the teacher chooses to ask it, a model hosted in the UK is shown the numbers from the saved versions and the largest additions,with no name, email address or college, and what it answers is advice to the teacher that nothing acts on. Section 5 has the rest.

4. Code You Run in the Python Playground

Code you run is sent to a separate service that executes it in an isolated, short-lived container, and is thrown away with the container. What we keep is the file you saved and the result of the last test run against it.

Please do not put personal data about other people, or confidential material belonging to an employer or a client, into a file or a playground session. Nothing there needs it, and clause 3 of our Terms & Conditions asks you not to.

5. Ask AI

The learning platform has one AI helper, Ask AI, in three places. A student can use it on a coding task where their teacher has turned it on, to ask questions about the task and their own code. A member of staff can use it to mark a student’s handed-in work: to review it, suggest a mark and write feedback. And staff can use it to build a task, which is described below. Nothing is sent anywhere until somebody types a question and sends it.

For a student, what goes to the model is what they type, the task’s description and the code they last saved, and nothing else: no name, no email address, and nothing that identifies whose code it is. It is told to help them think rather than to give the answer, it cannot change anything on their screen, and it only talks about the task: asked about anything else it says it cannot answer. Before a student sees an answer a second model checks it, and one that is not about the task or that gives the answer away is not shown. When a teacher marks with it, what goes is the task’s description, the student’s saved code, the mark and feedback in the box, and what the teacher types; what comes back is a suggestion the teacher reads, and nothing reaches the student until the teacher saves it.

Everything that touches a student’s work, their own questions and a teacher marking it alike, runs on Amazon Bedrock, in the AWS London region, the same region the rest of the platform’s data is held in, so it does not leave the United Kingdom. AWS acts as our processor for it, see section 6, and under the AWS service terms material sent to Bedrock is not stored by the model service, is not used to train models, and is not shared with the companies that make them.

The third use is for staff only, and never for students: a teacher can have a quiz or an assignment drafted from what they type and from material they choose, which is files they have already uploaded to their own module and any web pages they paste in. What goes to the model is that material and their instruction, and nothing else: no name, no email address and no student’s work of any kind. Nothing is created until the teacher reads the draft and presses Create.

That one can run on Amazon Bedrock in London, on Google’s Gemini API or on OpenAI, and which of them is a choice a site administrator makes model by model. Google and OpenAI act as our processors for it, see section 6; under both of their paid API terms, what is sent is not used to train their models. Those two are the only parts of the platform whose processing is not confined to the United Kingdom, the Bedrock option stays in London like everything else, and the page an administrator chooses on says which is which. All of it is switched off unless a college turns AI on for its staff, and no student’s work is ever sent to any of them.

What we keep afterwards is a count, how many questions a student has asked on each task and how much each month’s calls cost, so the allowances can be enforced and the college billed accurately, and the conversationitself: what was typed and what was shown in reply, so somebody can carry on where they left off. A member of staff’s conversation is theirs alone. A student’s can be read by a teacher who may already see that student’s work on the task, and the student is told so in the Ask AI panel itself before they ask anything; each time a teacher reads one, it is written to the student’s access log, which the student sees. What a student asked before that notice was added on 5 October 2026 was asked in private and is never shown to anybody. A conversation is one per person per task, or for a teacher marking, one per student’s piece of work; it is emptied whenever its owner presses New conversation and deleted 90 days after each line was written. The code and material it read are not copied into it. Ask AI can be switched off for a whole college, for staff or students separately, or for a single task, and where it is off nothing is ever sent.

6. Who Else Sees It

Processors: they act on our instructions

We are answerable for these, and each is under a contract that says what it may do.

Third parties acting as their own controller

Google, for advertising measurement (section 9), decides its own purposes and does not act on our instructions, so it is a controller in its own right, and we and Google are jointly responsible for one part of it: the tag on our pages collecting that data and sending it to Google. It only happens if you accept the banner. What Google does with that data is set out in its privacy policy and at policies.google.com/technologies/ads, and rights over it can be exercised with Google directly as well as with us.

YouTube, for a video a teacher has put in a lesson (section 9), is part of Google and is a controller in its own right in the same way: it decides what its player does and what plays in it. Unlike the advertising measurement above, it is not something the banner turns off, because the video is part of the lesson the teacher chose. What YouTube does with what it receives is set out in Google’s privacy policy, and rights over it can be exercised with Google directly.

The other sites a teacher can show in a lesson (section 9), Vimeo, Loom, Figma, GeoGebra, Desmos, PhET, Google Docs, Slides and Sheets, and Microsoft 365, are each a controller in their own right in the same way, for the same reason: each decides what its own page does once it is loaded. Rights over what any of them receives can be exercised with that company directly.

And otherwise

Your college, where you are its student and the work is theirs to see; our accountant and, if it ever came to it, our insurers or solicitors; and anybody we are required by law to tell. If the business were ever sold, records would pass to the buyer, who would be bound by this policy.

Requests from public authorities

If the police, a court, a regulator or another public authority asks us for information about you, we do not hand it over on a telephone call or an informal email. We ask for the request in writing, saying who is asking, whose information they want and the law that lets them ask, and the director of the company decides what is done. We check that it is lawful and no wider than it needs to be, take legal advice where it is unclear, and ask the authority to narrow it, or challenge it, where it is not. We give only what is asked for and nothing more. Unless the law or an order forbids it, we tell you before we do, or as soon as we can, so that you can object. If you are a student or a member of staff at a college and the information is your college’s, we send the authority to the college, which decides, and we tell the college. We do not answer an authority outside the United Kingdom directly, only one whose request has come through UK legal process. We keep a record of every request and what we did about it, and the procedure is reviewed every year.

7. Where Your Data Is Held

In the United Kingdom. The site, the database and every saved file are in the AWS London region and do not leave it in the ordinary course of things. Git repositories are on a server we run on our own machine in the United Kingdom, copied every night to the same London region. Some of the services above are reachable from outside the UK, and this is what each transfer relies on:

This is the position as at the date at the top of this page; these mechanisms change when the law does, and we update the table when they do. For a copy of the clauses that apply to any of them, email [email protected] and we will send you the relevant addendum, or a link to the provider’s published version of it.

8. How Long We Keep It

• Enquiries that go nowhere: up to 12 months, then deleted.
• A competition registration: the address you entered is recorded, with the country, city and time zone Cloudflare works out for your connection (never your IP address), and with whether you opened the link we emailed, so we can send it again if you try to sign in first and can tell who registered and who confirmed. If you never open the link, the address is deleted 365 days after you registered. Once you do, the account is the record.
• A sign-up you started and did not finish: 30 days from the last time you asked for a code, then the address is deleted whether or not anybody has acted on it. It is deleted at once if you finish and the account is created, because at that point the account is the record.
• Customer records, bookings and payments: 6 years after our last dealing with you, which is what HMRC and the Limitation Act between them require of the records behind an invoice.
• An Ask AI conversation, a student’s or a member of staff’s: 90 days from when each line was written, or sooner if they press New conversation. See section 5.
• Coaching notes, from when we offered career coaching: deleted within 12 months of the last session unless you ask us to keep them.
• An Ubuntu VM kept by a free account: deleted with everything on its disk once nobody has started it for 90 days. We email 7 days before, and starting it once keeps it. A VM started without an account keeps nothing after it stops.
• Student work on the platform: for as long as the college says, and deleted on its instruction.
• Git repositories: the same, including after an account has been removed, until the college or you ask us to delete them. The disk they are on is copied every night and each copy is kept for 30 days, so a repository deleted today is gone from those copies 30 days later, its large files and attachments with it.
• An account you have asked us to delete: we stop you signing in straight away and delete the account 30 days later. The 30 days are so you can change your mind, and there are two ways to: the link in the email we send you when you ask, or by asking us. After that it is deleted, and what survives is only what we are separately required to keep, listed above, above all the records behind an invoice you have already paid.
• Access and administrative logs: these are append-only records of who did what, no row is ever edited or removed on its own, and they are what lets a student see who has opened their work. We keep them for the life of the account and 12 months after it is closed, then delete them. We review what is held once a year, and delete anything that no longer needs to be.
• Ask AI: a count of questions asked on each task and of each month’s cost, kept with the account’s other records. The conversation itself is covered by its own line above, see section 5.
• Marketing consent: until you withdraw it, plus a note that you have withdrawn it, so we do not write to you again by mistake.

9. Cookies and Similar Technologies

Nothing beyond the entries below is stored on your device unless you agree to it. The first time you visit, a banner asks. Until you answer, and if you answer no, this site is exactly as it is described in the next paragraph and nothing is loaded from Google at all.

Strictly necessary, and set without consent

None of these needs your consent: the first is what signing in is, the second, the ones that remember how you want to read and the one that remembers your language exist only to record a choice you made, and the last is how a page you asked for shows you the right time. All are first-party, and none is used to track you.

The security check on our forms. Cloudflare’s Turnstile, which is what asks whether you are a robot before a form is accepted, may set a short-lived cookie of its own while it does so. It is there to carry the result of that one check, it is not used for advertising or to follow you anywhere, and without it the forms cannot tell a person from a script. That makes it strictly necessary too, so it is not behind the banner. Cloudflare’s account of what Turnstile collects is in its privacy policy.

Google advertising measurement, only if you accept

If you press Accept, the page loads Google’s tag (gtag.js, account AW-18079244030, from www.googletagmanager.com). Google uses it to measure whether somebody who saw one of our adverts went on to enquire. It sets cookies of its own, whose names begin _gcl or _ga, and it tells Google that a page was loaded.

We also tell Google when somebody sends the college form at the end of signing an organisation up, which is the one point in the site where we report a conversion rather than a page view. What is sent is that it happened, not who did it or anything typed into the form: no name, no address, no email address. It is sent only if you accepted, and if you did not, the form is sent to us and nothing at all goes to Google.

It is reported at that step, and not once the sign-up has finished, for a reason worth stating: the page you land on afterwards is inside the teaching platform, and we load no advertising there at all — whatever you accepted — because the people using it include students. That is the promise in the next paragraph, and it costs us some accuracy in this one.

For this measurement Google decides its own purposes and is not acting on our instructions, so it is a controller in its own right rather than our processor, see section 6. What it does with advertising data is set out at policies.google.com/technologies/adsand in its privacy policy.

If you press Reject, or close the banner without answering, the tag is never requested. It is not blocked in your browser after loading: it is not put on the page in the first place.

It also never loads inside the learning platform, whatever anybody has accepted. Once you are signed in to a college account, no advertising tag appears on any page: students are often under 18, and coursework is not something to measure adverts against. See section 13.

And it never loads on the pages written for students, signed in or not: the free lessons, the Resources pages, and the competition pages, which are advertised to school students and whose own form asks for an age band and takes nobody under 16. Those carry no advertising tag whatever anybody has accepted, which costs us the ability to measure anything on them and is the right way round.

Changing your mind

Cookie settings at the foot of every page reopens the banner, and either answer can be given as often as you like. Refusing after having accepted also clears the _gcl and _ga cookies and reloads the page, because a tag already running cannot be stopped in place. We ask again every 6 months, so a yes does not last forever and neither does a no.

What else your browser fetches

The code editor is fetched from a public CDN (cdn.jsdelivr.net) on the pages that carry one, because it is far too large to serve ourselves. The styling and everything else we write comes from us. Those requests set no cookies, but like any request they tell that CDN your IP address. They are how the page is delivered rather than a choice about you, so they are not something the banner can turn off.

A lesson can include a YouTube video, if the teacher who wrote it added one. On a lesson that has one, your browser loads YouTube’s video player from www.youtube-nocookie.com as the page opens, before you press play. That request tells YouTube, which is part of Google, your IP address and which video was loaded on which of our pages. It happens only on a lesson that holds a video, and not anywhere else on the site.

We use YouTube’s privacy-enhanced mode, which is what that address is. YouTube says that in this mode it does not store information about you until you play the video, and that a video watched this way does not shape what it recommends to you. If you press play, YouTube may set its own cookies and record what you watched, under Google’s privacy policy.

YouTube decides what plays inside its player, and on some videos that includes YouTube’s own adverts. We do not choose them, cannot turn them off from here, and are sent nothing back. For this Google is a controller in its own right and not our processor, see section 6. Because the video is part of what the teacher chose to put in the lesson, it is not something the cookie banner controls.

A lesson can also show a page from another site, if the teacher who wrote it added one: a video from Vimeo or Loom, a design from Figma, a graph or construction from GeoGebra or Desmos, a simulation from PhET (the University of Colorado Boulder), or a document from Google Docs, Slides or Sheets or Microsoft 365 (OneDrive and SharePoint). Your browser loads it from that site as the lesson opens, which tells that site your IP address and which of our pages it was shown on, and that site may then set its own cookies under its own privacy policy. It happens only on a lesson that holds such a page, and not anywhere else on the site. Vimeo videos are loaded with Vimeo’s “do not track” setting on. A Google or Microsoft document can be shown only if its owner has made it readable by anybody with its address, and if it asks you to sign in, that is the provider asking, not us.

We run no other analytics: no Google Analytics, no pixel from any other advertising network, no session recording, no heat maps, and no cross-site profiling by us.

10. Your Rights

Under the UK GDPR you can ask us to:

• give you a copy of the personal data we hold about you;
• correct anything about it that is wrong;
• delete it, where we no longer need it and no law requires us to keep it;
• restrict what we do with it while a dispute about it is sorted out;
• send it on, to you or to somebody else, in a portable format;
• stop, where we rely on our legitimate interests and you object, or where we are using it for direct marketing, which we will stop on being asked, no reason needed.

If you booked a call with us, or you signed up for yourself rather than through a college, you can delete your own account, from the Your account tab on your profile page. It does not happen at once: we stop the account signing in immediately and delete it 30 days later, so that a deletion nobody meant can be undone. You can undo it yourself from the link we email you when you ask, or by asking us, at any point in those 30 days. Bookings you have already paid for stay, because the records behind an invoice are ones we have to keep, and section 8 says for how long.

If you are a student or a member of staff at a college, ask the college.Your account, your work and your marks belong to a course the college is running, and the college rather than us decides what happens to them, which is why the button on your profile page says Request to delete rather than deleting it. Pressing it tells whoever administers the platform at your college, and they can delete your account for you; nothing about your account changes in the meantime. You can tell them directly instead, and if you would rather come to us, write to us at the address below and we will pass it on and see it is dealt with. It is worth saying that this right is not absolute for anybody: where a college has to keep something for an awarding body or for an appeal, that is one of the reasons the law allows for keeping it, and the college will tell you if it applies to you.

Where we rely on your consent, you can withdraw it at any time, and withdrawing it does not make what we did beforehand unlawful. Email [email protected] for any of these. It costs nothing, and we may ask you to confirm who you are first, so that somebody else cannot ask for your data.

We answer within one month. The law allows up to three months for a request that is genuinely complex or where somebody has made several at once; if that ever applied to yours we would tell you inside the first month, and say why. We would rather not need it.

11. Whether You Have to Give Us Anything

No, but some of it is the service. We cannot confirm a booking without an email address to send the confirmation to, and we cannot take a payment without the payment going through a provider. Marketing consent is genuinely optional, and refusing it changes nothing else. Nothing in a coaching conversation has to be answered.

12. Automated Decisions and Profiling

We do not make decisions about you by automated means, and we do not profile you. A person reads every enquiry and a person takes every call. The site does score submitted forms for signs of automation, in order to reject spam, which decides whether a form is accepted rather than anything about you: if a form of yours is refused, write to us and a person will deal with it.

Ask AI in section 5 decides nothing about anybody and marks nothing: an answer is something a student reads, a suggested mark or piece of feedback is something a teacher reads, and a person decides what to do with both. A suggested mark only reaches a student if a teacher saves it.

13. Children

Our services are sold to adults: clause 5 of the Terms & Conditions requires anybody booking or paying to be 18 or over, and we do not knowingly take a booking from a child.

Learning platform accounts are a different case, because college students are often 16 or 17. A college account is created by the college, not by the student, and the college is the controller of that data and responsible for the consents and the safeguarding around it. We hold no more than the platform needs, we do not market to any account, and we do not use student work for any purpose of our own.

Free lessons on our Resources pages work differently. Anybody can read one without an account at all. Sending answers to be marked needs an email address, and the account that creates is made by the person using it rather than by a college, so we are the controller for it and there is no college behind it. For those accounts we hold the email address, the answers sent, the marks, the dates, the time zone your browser reported so the dates read correctly, and whether you have asked us to stop emailing you about your work: no name is asked for and none is needed. No advertising or analytics script loads on a free lesson page, on the Resources pages or on the competition pages, whatever anybody has accepted elsewhere on the site, because these are the pages most likely to be read by somebody under 18.

Registering for a competition asks for little, and says so on the form. Registering for one of our coding competitions creates the same kind of self-service account. It asks for an email address and nothing else. If you later choose to tell us more, we can hold four more answers: a name or a nickname of your choosing, the university or college where you study, an age band, and your time zone. We ask for a band rather than a date of birth deliberately, because the band answers everything the competition needs and is the lighter thing to hold about somebody who may be as young as 16. Entrants must be 16 or over, and anybody under 18 who wins needs a parent or guardian to agree to the prize before it is paid. Proof of enrolment is checked for winners only, before a prize is confirmed. The competition page states all of this in its terms before you register.

If you believe we hold data about a child that we should not, or you want a free-lesson account and its answers deleted, write to [email protected] and we will look into it and delete it if that is right. Deleting one of those accounts costs nobody anything: there is no course and no qualification attached to it.

Because the platform and the free lessons are both likely to be used by 16 and 17 year olds, we build them with the ICO’s Age Appropriate Design Code in mind, even where the college is the controller. In practice that means: no marketing to a student or free-lesson account, ever; no profiling and no automated decisions about students; no advertising or analytics of our own inside the signed-in platform or on any free lesson page, with the one exception that a teacher may put a YouTube video in a lesson, which loads YouTube’s player and can show YouTube’s own adverts, or a page from one of the other sites section 9 names, which loads that site’s page, as section 9 sets out; the least data the platform can work with, which for a free lesson is an address and the answers and nothing more; and every staff look at a student’s work recorded in a log the student can read in full themselves, which is there so that being watched is never invisible to them.

14. Changes to This Policy

We update this policy when what we do changes. The date at the top is the version you are reading, and a change that materially affects you will be sent to you by email rather than left here for you to notice.

15. How to Contact Us, and How to Complain

• by email: [email protected]
• by post: Mitsuru Technologies Ltd, 11 Poppy Mills Road, Brockworth, Gloucester, GL3 5AB, United Kingdom
• online: our contact form

Please come to us first, because most of this is quicker to fix than to complain about. You have the right to complain to the Information Commissioner’s Office at any time, whether or not you have asked us: ico.org.uk, or 0303 123 1113.

Mitsuru Technologies Ltd, registered in England & Wales, company number 16755264, VAT number GB506417704. Registered office: 11 Poppy Mills Road, Brockworth, Gloucester, GL3 5AB, United Kingdom. Terms & Conditions. Last updated 15 October 2026.