Privacy Policy
Last updated: · version 1.6
This policy describes what the Site and the application actually do with your data as of the date above, including the audience measurement and the service-usage statistics, which have run since 11 October 2026 and which you can object to at any time (see “Audience measurement” and “Service-usage statistics”).
This English version is provided for convenience. Only the French version is legally binding.
Who is responsible for your data
The data controller is Mathieu Soysal, publisher of the Site (see the legal notice), reachable at contact@teacher-and-me.com. No data protection officer has been appointed.
This policy covers the website teacher-and-me.com, the application reached from it and the public pages it generates (blog, level test, path catalogue). It describes what the Site actually does with your data as of the date at the top of the page.
Tutors, students and who is responsible for what
The Site has two kinds of users: tutors (independent teachers), who create paths of exercises, and students, who follow them. Everyone creates their own account on the same sign-in page, except a student who has no e-mail address: their tutor can create an account for them (see “Learner accounts created by a tutor”). The role is chosen at sign-in and can be changed at any time, except for those accounts, which stay in student mode.
For accounts, sign-in, security, audience measurement, the service-usage statistics and the general running of the Site, the publisher is the data controller.
A tutor who enrols a student on a path, writes exercises, reads their students’ progress or sends them messages decides why and how that data is used in their own teaching. For that teaching data the tutor is the controller and Teacher and Me processes it on the tutor’s behalf, as a processor, for the sole purpose of providing the service. A written data-processing agreement (Article 28 GDPR) is available on request at contact@teacher-and-me.com. The tutor is answerable for the lawfulness of the data they enter about their students, including the parental authorisation required for students under 15 (see “Minors”).
The data we process
Account
E-mail address, display name (by default the part of your e-mail address before the @; editable in Settings), role (tutor or student), the “hide my profile” option, the account’s interface language, creation date and a technical id. We record no date of birth, phone number, postal address or photo.
Five more items serve the service-usage statistics (see that section): the role chosen when the account was created, the date you first switched to teacher mode, your answer to “How did you hear about Teacher and Me?” if you gave one, the channel code if you agreed to keep it when creating your account, and your choice to leave your account out of those statistics or not.
Learner accounts created by a tutor
A tutor can create an account for a student who has no e-mail address. It holds the first name or nickname the tutor typed and the interface language, then the student’s progress and messages; it has no e-mail address and no password. It opens with a personal link, also shown as a QR code, which the tutor gives the student; we keep that link hashed, to recognise it, and encrypted, so the tutor can show it again, with the date it was last used. The tutor can replace the link, which stops the old one working, and delete the account. Deleting the tutor’s account deletes the learner accounts they created. These accounts are never read by the service-usage statistics.
Sign-in and security
- E-mail code: a 6-digit code, stored hashed, valid 10 minutes. On the very first sign-in with an unknown address, the account is created immediately for that address, without first verifying the mailbox; code verification applies from the next sign-in on. Only use an address you own.
- Google account: Google sends us your e-mail address and display name; we keep no other data from your Google account.
- Passkey: credential id, public key, signature counter, transport type, creation and last-use dates. The private key never leaves your device.
- Personal link (learner accounts created by a tutor): opening the link opens a session; the device keeps a fingerprint of the link to open it again.
- Session: an access token valid 15 minutes, kept in memory, and a single-use refresh token valid 7 days, stored in your browser’s local storage and, hashed, on our servers. No screen of the app asks for a password. Our API still accepts, though no screen uses it, signing up and signing in with a password: for an account created that way, the password is stored hashed.
- No IP address is stored in our databases. IP addresses are used transiently by Cloudflare to rate-limit requests and protect the Site against abuse, and appear in its technical logs (see “How long we keep data”); our audience-measurement service uses them in memory only (see “Audience measurement”).
Teaching content (tutors)
Paths, lessons and exercises (questions, choices, expected answers, explanations, optional YouTube link), audio clips you record or upload (stored inside the exercise itself), the list of your enrolled students, messages sent, paths published in the public catalogue.
Learning activity (students)
Exercises completed, position in the path, experience points, activity streak, time spent, number of attempts, answers typed (including wrong answers and the transcripts of pronunciation exercises), skipped exercises, a timestamped attempt log, the availability you declare for bookable lessons, your exam goal (exam, date, target level). The tutor of a path sees all of this for the students enrolled on that path; you can erase your progress at any time, and your tutor can erase your progress on their paths. Joining a path with the code or link its tutor shared enrols you on it: that tutor then sees your name, your e-mail address and your activity on that path.
Pronunciation exercises
Your microphone audio is transcribed in your browser; the recording is never sent to our servers. Only the transcript and the score are sent. On first use your browser downloads the speech-recognition model from Hugging Face’s and jsDelivr’s servers, which then see your IP address.
Messages
Tutors can send messages (congratulations, encouragement) to the students enrolled on their paths: content, sent date, read date. There is no student-to-tutor messaging. Each party can delete their own messages; a tutor can also delete the messages of a student they teach.
Public path catalogue
When a tutor publishes a path in the public catalogue, its title, description, README, language, level, skills and audience are shown publicly, with the tutor’s name unless they enabled “hide my profile”. Reviews (stars and text) are shown without a name, as a verified acquirer; comments are shown with the name chosen when posting; reports (reporter, reason, note) stay internal. The catalogue’s public pages are regenerated at every update of the Site and once a day, so a removal reaches them within 24 hours.
Welcome questionnaire (tutors)
When you first set up the tutor area, three answers (role, number of students, goal) are kept in your browser. They tailor the onboarding texts and the sample path, which is saved on our servers like your other paths; these three answers never leave your browser. The last step also offers an optional question, “How did you hear about Teacher and Me?”: if you answer it, your answer is kept with your account (see “Service-usage statistics”).
Technical data
Cloudflare keeps technical logs of a share of requests (one in 5): IP address, requested URL, request and response headers and metadata. Our error logs contain no name or e-mail address.
Why, and on what legal basis
The legal bases below are those of Article 6 of the General Data Protection Regulation (GDPR).
| Purpose | Data | Legal basis |
|---|---|---|
| Providing the service: accounts, paths, progress, messages, catalogue, offline use | account, teaching content, learning activity, messages, browser storage | performance of the contract (the terms of service); for teaching data, the instructions of the tutor acting as controller |
| Securing accounts and the Site: codes, tokens, passkeys, rate limiting, logs | sign-in data, IP address (transient, logs) | legitimate interest in the security of the service |
| Sending service e-mails: the sign-in code and, for a sign-up with a password, address verification | e-mail address | performance of the contract |
| Teaching assistant (tutors) | your questions, the anonymised context of your paths and students | performance of the contract, at your request |
| Measuring the audience of the Site and the app, to know which pages are read and how people move between them, find display problems and improve the Site | page types viewed, scrolling and time in bands, clicks on a fixed list of buttons and links, loading speed, the context of a visit’s first page, the daily visit identifier (see “Audience measurement”) | legitimate interest (Article 6(1)(f) GDPR); you can object at any time |
| Computing service-usage statistics, to understand how the service is used and improve it | the account data described in “Service-usage statistics”; only totals are kept | legitimate interest (Article 6(1)(f) GDPR); you can leave your account out at any time |
| Keeping with your account the channel code you agreed to keep when creating it (tutors) | channel code | your consent, given by ticking the box (Article 6(1)(a) GDPR), which you withdraw by removing the code in Settings |
| Keeping your answer to “How did you hear about Teacher and Me?” (tutors) | your answer, from a fixed list | legitimate interest in knowing which channels bring tutors to the service; answering is optional, and the answer can be removed at any time |
| Handling your feedback and automatic error reports | comment, account e-mail address, technical data you tick | legitimate interest in fixing and improving the service |
| Answering your requests and honouring your rights | your request, your e-mail address | legal obligation |
No data is sold, rented or used for advertising. We send no newsletter and no marketing e-mail.
Teaching assistant (artificial intelligence)
Tutors have an assistant that answers their questions about a student’s or a path’s results and can suggest changes to a path, applied only if the tutor approves them. Students have no access to it. It runs on Cloudflare Workers AI, with the @cf/mistralai/mistral-small-3.1-24b-instruct model and, beyond a monthly quota or when that model is unavailable, @cf/meta/llama-3.1-8b-instruct-fast.
- What is sent: your question and a context (at most 16,000 characters) built in your browser from the path’s titles and content, statistics and wrong answers. Your students’ names are replaced with “Student A”, “Student B” and so on before anything is sent; the application reminds you of this every time.
- What is kept: the context is neither stored nor logged by our servers. The question and the answer of each exchange are kept, limited to the last 24 messages per discussion, until you clear them (“New conversation”). Usage counters (token counts) are kept per tutor.
- Where: discussions are stored in a European region of Cloudflare; the models run on Cloudflare’s global network.
- Use by the provider: under its published terms, Cloudflare does not use this content to train models or improve its services, and does not make it available to any other customer.
- No decision with legal or similarly significant effects is made automatically: the assistant suggests, the tutor decides.
Audience measurement
Since 11 October 2026, the Site and the application have measured their audience with our own programs, hosted by Cloudflare. The measurement tells us which pages are read and how people move between them, shows display and navigation problems, and helps us improve the Site. It is not used for advertising or to build a profile, and no decision about you is based on it. You can refuse it at any time, without an account and with no effect on how you use the Site (see “Refusing audience measurement”).
It covers the homepages, the guides, the pricing page, the accessibility statement, the blog, the level test and the application. It does not cover the legal pages (this one included), the press kit, the error pages or the Site’s administration pages.
What your browser sends
- Page views: the page type, that is the pattern of its address without any identifier (for example
/paths/:pathId), or the published address of a page of the Site. Nothing after a “?” or a “#” in the address is sent. - Scrolling and time: for each page view, how far down you scrolled (0, 25, 50, 75 or 100%) and how long the page stayed on screen, in bands (under 10 seconds, 10 seconds to 30 seconds, 30 seconds to 1 minute, 1 minute to 3 minutes, 3 minutes to 10 minutes, 10 minutes or more).
- Clicks: on a fixed list of buttons and links, named by a technical label, and on the links you follow, named by their destination (the page type on the Site, the site’s name for another site, the file type for a download). The element’s text and anything you type are never sent; links written by users (a catalogue path’s description, an assistant answer) are not measured.
- Loading speed: for one page load in 4, drawn at random, four indicators (LCP, INP, CLS, TTFB), each rated good, needs improvement or poor.
- The context of a page view: the site you came from, reduced to its domain name and only when it is another site, for the first page; the page’s language; the device category (phone, tablet or computer), worked out from the window’s width; and whether the app is installed on the device.
Your browser sends these without attaching its cookies, and they never include your account id, your name, your e-mail address or your answers. For the first page of a visit, our server adds the country and region as Cloudflare works them out from your IP address (never the city), the name and major version of your browser and the family of your operating system.
In student mode, a page view carries the page type only: no context is sent and our server adds none. Scrolling, time, clicks and loading speed are measured as for everyone.
The daily visit identifier
To count visits without storing anything on your device, our server computes, for each request it accepts, a pseudonymous daily visit identifier: a fingerprint of your shortened IP address (its first 3 bytes for IPv4, its first 48 bits for IPv6), the string your browser identifies itself with (“User-Agent”), a secret key of the Site and a key drawn at random for each day (in universal time). Your IP address and that string are never stored.
The identifier is used only to group the pages of one visit (a visit ends after 30 minutes without a page view and never runs past the day), to count the day’s visitors and to turn away abusive traffic (more than 120 requests a minute). The day’s key is deleted at most 27 hours after its day starts, and the visit records (identifier, visit number, start and end times, first and last page, number of pages, steps reached in the paths described below) at most 51 hours after. While the day’s key exists, someone who also held the Site’s secret key could recompute the identifier of a browser whose IP address and identification string they knew.
What we compute
Every night a program computes daily totals from them: page views per page type, visits and visitors, entry and exit pages, moves from one page to another, scrolling, time, clicks and loading speed, and, for the first page of each visit, the breakdown by referring site, country and region, language, device, browser, operating system and installed app. The number of visits per entry page and referring site is the only combination of these.
It also counts how many visits get through the steps of four paths: from the sign-in page to the app; from a homepage to the sign-in page, then to the app; from a blog post to the sign-in page, then to the app; from the level test’s page to its start button, then to the sign-in page and the app. For the last two, only the posts and the test in English and French are counted.
Daily totals are kept 35 days; weekly (Monday to Sunday) and monthly totals, rounded to the nearest ten, 25 months, and a value seen fewer than 10 times in the period is grouped under “other”. Requests from bots and malformed requests are turned away: only their number is counted.
The statistics are kept in a Cloudflare database restricted to its EU jurisdiction, and so in the European Union. The requests are processed in memory by the Cloudflare data centre that receives them, and the service that receives them keeps no technical log.
Why you see no consent banner
The audience measurement sets no measurement cookie and writes nothing on your device to measure; it only produces the statistics described above, which only the publisher reads. We run it on the basis of our legitimate interest in knowing and improving how the Site is used (Article 6(1)(f) GDPR), without asking for your consent, and you can object to it at any time, as described below.
Reading or writing information on your device is governed by Article 82 of the French Data Protection Act (loi Informatique et Libertés) (in French). The CNIL, the French data protection authority, sets out on its page on audience measurement (in French) the conditions under which audience measurement can run without consent, and we designed ours to meet them. That assessment is our own: the CNIL has not reviewed our measurement.
Refusing audience measurement
You can refuse audience measurement at any time with this button or, in the app, in Settings, under “Usage statistics”.
Your refusal is kept in a cookie, tam_audience_optout, which our server sets for teacher-and-me.com and its subdomains. It holds only your refusal and is kept 400 days, renewed each time you open the refusal page, which also shows your current choice and lets you withdraw it. While it is there, neither the Site’s pages nor the app send any measurement. It applies to this browser only: in another browser or on another device you need to refuse again, and clearing the site data in your browser clears it. On iPhone and iPad, the app installed on the Home Screen has its own data: refuse there too, in its Settings or on this page.
If your browser sends the Global Privacy Control or Do Not Track signal, we treat it as an objection: your browser sends nothing, and our server discards any request carrying the signal before reading it. Your refusal covers the audience measurement only: it changes nothing in Cloudflare’s technical logs and protections, nor in the service-usage statistics, which you can leave separately (see “Service-usage statistics”).
Until 11 October 2026: PostHog
Until 11 October 2026 the Site and the application used PostHog (PostHog, Inc., on its European cloud in Frankfurt) to measure the audience, capture clicks automatically, record browsing sessions and produce heatmaps; once you were signed in, that data was tied to your account’s technical id, never to your name or e-mail address. Our PostHog project and all the data in it have since been deleted.
The ph_phc_uNHEKAjjPRnUcjLzTWRka6QiYonjCTW3E7AD4E3YCtHb_posthog cookie and the storage keys PostHog left in your browser are deleted on every visit, by the application, the level test and the measurement script of the Site’s pages, even if you have refused audience measurement, until at least 11 November 2027. Nothing is stored to remember that the deletion happened.
Service-usage statistics
Since 11 October 2026, every night we compute service-usage statistics from account data, to understand how the service is used and improve it. They are totals only: how many accounts are created each week (Monday to Sunday), by the role chosen when creating them; how many of the tutors who arrived in a given month create their own first path within 7, 30 and 90 days, or start from the sample path, a catalogue path or an adopted programme; how many of them open the app each week; and, for each channel code and each answer to “How did you hear about Teacher and Me?”, how many tutors they account for and how many of them create their own first path within 30 days. A tutor “arrives” in the month they first switch to teacher mode.
What the computation reads
For each counted account: its technical id, creation date and role, the role chosen when it was created, the date it first switched to teacher mode, its answer to “How did you hear about Teacher and Me?” and its channel code. For each counted tutor: the dates their sessions were opened or renewed in the week just ended, and the creation date and origin of their paths (created, sample, catalogue or adopted programme), without their title or content. The computation reads no e-mail address, no name, no student’s progress, answers or messages, and no audience-measurement data.
Accounts left out of the statistics, learner accounts created by a tutor and test accounts are never read. A student account counts in one figure only: the number of accounts created that week, by role. What is read is matched in memory, and only totals are written.
The totals
Each figure counts distinct people. A figure that would count 1 to 4 people is never stored, and a share is stored only if at least 5 people reached it and at least 5 did not; codes and answers that are too rare are grouped until they count at least 5 people. Each figure is computed once and never recomputed. The totals are kept 25 months after their week or month, in the same database as the audience measurement, restricted to the European Union, but in separate tables: no figure combines the two. Only the publisher reads them, in particular on an administration page open only to the accounts the publisher designates.
Leaving the statistics
In the app’s Settings, under “Usage statistics”, the box “Leave my account out of usage statistics” takes your account out of all these figures from the next nightly computation, on all your devices; figures already computed are not recomputed. If you untick it, your account is counted again from the next computation. You can also ask us by e-mail (see “Your rights”). This box is separate from refusing audience measurement, which applies to one browser.
The channel code
A link to a homepage or to the Site’s sign-in page may carry a channel code, which names a channel through which people find the service; the list of active codes is public. The homepage carries the code over to its links to the sign-in page, storing nothing. In teacher mode, the sign-in page then offers you an unticked box to keep it with your account. If you tick it and this sign-in creates your account, the code is kept with it; otherwise it is ignored. With a Google sign-in, the ticked code is set aside for at most 30 minutes, tied to no account, until you come back, then deleted, at the latest the following night.
You can remove the code at any time in Settings, under “How you found us”: it is deleted from your account and left out of every figure computed afterwards; figures already computed are not recomputed.
“How did you hear about Teacher and Me?”
This optional question is asked of tutors at the last step of the welcome questionnaire, then once on their dashboard if they have not answered it. The answer is one of 8 options, never free text, and is kept with your account. You can change it in teacher mode and remove it at any time in Settings, under “How you found us”; figures already computed are not recomputed. The app only notes, on your device, that the question was shown to you.
Feedback, bug reports and automatic reports
The application’s feedback button (and, for tutors, the automatic report sent when the assistant fails) records your comment, the date, the type of feedback, your account’s e-mail address, attached automatically when you are signed in, and, only if you tick the corresponding boxes, your screen size and browser description. This record is kept 14 days in a Cloudflare storage space.
Your comment is then rewritten by an AI model (@cf/mistralai/mistral-small-3.1-24b-instruct) instructed to remove personal information and posted as a ticket in a private GitHub repository. If that automatic anonymisation fails, the comment is posted as is. Your e-mail address is never sent to GitHub.
E-mails
We only send service e-mails: the sign-in code. Our API can also send an address-verification e-mail, or an e-mail saying the address already has an account, to someone who signs up with a password through the endpoint no screen of the app uses (see “Sign-in and security”). They are sent from no-reply@teacher-and-me.com through Cloudflare Email Service. There is no newsletter and no marketing e-mail. The Site sends no invitation e-mail: a tutor who wants a student to join a path shares its code or link themselves, for example from their own mail client. The “e-mail me my results” form of the level test currently sends no message and does not transmit the address entered.
Recipients and processors
Your data is shared only with the providers below, which either process it on our behalf (processors) or are contacted directly by your browser at your request (independent third parties, responsible for their own processing).
| Provider | Role | Functions | Data location | Transfers outside the EU |
|---|---|---|---|---|
| Cloudflare, Inc. 101 Townsend St, San Francisco, CA 94107 European entity: Cloudflare France SAS, 6 place de la Madeleine, 75008 Paris privacy policy | Processor | hosting of the Site and its API; databases and file storage; running the AI models (assistant, anonymisation of feedback); sending the sign-in e-mails; technical logs and abuse protection; audience measurement and service-usage statistics | databases in Europe (EU or UK), with those of the assistant discussions and the statistics restricted to the European Union (details below); global network for delivery, security, AI and receiving the audience measurement | EU-U.S. Data Privacy Framework certification and European Commission standard contractual clauses |
| GitHub, Inc. 88 Colin P. Kelly Jr. St., San Francisco, CA 94107 European entity: GitHub B.V., Prins Bernhardplein 200, 1097JB Amsterdam privacy policy | Processor | receiving anonymised feedback (private tickets) | United States | EU-U.S. Data Privacy Framework certification and European Commission standard contractual clauses |
| Google United States privacy policy | Independent third party | signing in with a Google account, at your request; thumbnails of the YouTube videos a tutor links | United States | not applicable: your browser contacts it directly, at your request |
| Hugging Face, Inc. United States European entity: Hugging Face SAS, 9 rue des Colonnes, 75002 Paris privacy policy | Independent third party | downloading the speech-recognition model | United States | not applicable: your browser contacts it directly, at your request |
| jsDelivr (cdn.jsdelivr.net) privacy policy | Independent third party | downloading the speech-recognition model | not stated by the provider | not applicable: your browser contacts it directly, at your request |
| Ko-fi Labs Limited Suite 501, The Nexus Building, Broadway, Letchworth Garden City, Herts, SG6 9BL (company no. 11087704) privacy policy | Independent third party | donations, on its own platform | United Kingdom | not applicable: your browser contacts it directly, at your request |
Where your data is hosted
Our databases are spread over the following Cloudflare regions (D1 and R2), as configured on the date above:
| Data | Region |
|---|---|
| Accounts and sign-in | Eastern Europe |
| Paths, exercises, enrolments, catalogue | Eastern Europe |
| Progress and exam goals | Western Europe |
| Messages | Eastern Europe |
| Assistant discussions | European Union (EU jurisdiction) |
| Feedback (files) | Western Europe |
| Audience measurement and service-usage statistics | European Union (EU jurisdiction) |
Cloudflare’s European regions are not limited to the European Union: a database placed in them can sit in the United Kingdom. Only the assistant discussions database and the database of the audience measurement and service-usage statistics are restricted to Cloudflare’s EU jurisdiction, which guarantees storage in the Union. Accounts, content, progress, messages and feedback are therefore stored in Europe (EU or UK), and the assistant discussions and the statistics in the European Union.
Until 22 October 2026, a read-only backup of the discussions made before 8 October 2026 is kept in the former database, located in a North American region of Cloudflare (United States); it will then be deleted.
AI models run on Cloudflare’s global network, with no choice of region. The audience-measurement requests and the computation of the service-usage statistics are processed in memory by Cloudflare’s data centres, and only their totals are written to the database above. Technical logs and abuse protection are processed by Cloudflare in its European and US data centres. The anonymised feedback tickets are hosted by GitHub in the United States.
For these transfers outside the European Union, Cloudflare and GitHub rely on their EU-U.S. Data Privacy Framework certification and on the European Commission’s standard contractual clauses, incorporated in their data-processing agreements (links in the table above).
How long we keep data
| Data | Retention |
|---|---|
| Account, teaching content, progress, messages, catalogue activity | for the life of the account; immediate deletion at your request (Settings → delete my account); no automatic deletion of inactive accounts; learner accounts created by a tutor are also deleted with that tutor’s account |
| Account items used for the service-usage statistics (role chosen at creation, first switch to teacher mode, choice to leave the statistics) | for the life of the account |
| Answer to “How did you hear about Teacher and Me?” and channel code | until you remove them, and at most for the life of the account |
| Channel code set aside during a Google sign-in | at most 30 minutes; deleted at the latest the following night |
| Audience measurement: the day’s key | at most 27 hours after its day starts |
| Audience measurement: visit records (with the daily visit identifier) | at most 51 hours after their day starts |
| Audience measurement: daily totals | 35 days |
| Audience measurement: weekly and monthly totals, rounded to the nearest ten | 25 months |
| Audience-measurement refusal cookie | 400 days, renewed each time you open the refusal page |
| Service-usage statistics (totals) | 25 months after their week or month |
| Technical restore points of the databases | 30 days |
| Sign-in codes | 10 minutes |
| Session tokens | access: 15 minutes; refresh: 7 days (expired tokens purged daily) |
| Assistant discussions | last 24 messages per discussion, until the tutor clears them or the account is deleted |
| Feedback (with e-mail address) | 14 days; the anonymised ticket is kept |
| Cloudflare technical logs | 7 days, for one request in 5 |
What the Site stores in your browser
The Site sets one cookie only, and only if you ask for it: tam_audience_optout, which keeps your refusal of the audience measurement. Nothing is stored in your browser to measure the audience. The application stores there what it needs to work, including offline, and your interface preferences; Cloudflare’s bot protection may add its own cookies, and Cloudflare asks your browser to report connection errors to it. The last column says why you are not asked for consent.
| Key | Type | Purpose | Lifetime | Why no consent |
|---|---|---|---|---|
tm.refreshToken | local storage | refresh token of your session (valid 7 days, single-use, replaced every time it is used) | until you sign out | essential to the service you ask for |
tm.lastUser | local storage | instant display of your profile at start-up (id, e-mail, name, role, “hide my profile” option, interface language, whether a tutor created the account, exclusion from the usage statistics) | until you sign out | essential to the service you ask for |
tm.lastRole | local storage | last role used, offered by default on the sign-in page | until you clear the site data in your browser | interface preference |
tm.lang | local storage | the language you chose for the app’s screens, so it opens in that language next time | until you clear the site data in your browser | interface preference |
tm.passkeyHint | local storage | records that a passkey exists on this device | until you clear the site data in your browser | essential to the service you ask for |
tm.passkeyPromptSeen | local storage | records that the offer to create a passkey was seen | until you clear the site data in your browser | interface preference |
tm.linkTokenHash, tm.linkLearner | local storage | records that this device opened a learner’s personal link, with a fingerprint of the link to open it again | until you clear the site data in your browser | essential to the service you ask for |
tm.teacherOnboarding.v1.<id> | local storage | teacher onboarding: steps seen and your three answers to the welcome questionnaire | until you clear the site data in your browser | essential to the service you ask for |
tm.heardAboutCardSeen | local storage | records that the question “How did you hear about Teacher and Me?” was shown to you (not your answer) | until you clear the site data in your browser | interface preference |
paths-v4, studentProgress-v4, messages-v1, studentDirectory-v1 | local storage | offline copies of your data: paths and exercises, progress, messages, your student directory | until the account is deleted (kept after a plain sign-out) | essential to the service you ask for |
pendingInvitations-v1 | local storage | offline copy of the e-mail invitations you had pending (path, address invited, date), kept by earlier versions of the app; the app no longer uses it | until the app next opens, which deletes it | no longer used; deleted the next time the app opens |
outbox-v1 | local storage | queue of changes made offline, sent when the network is back and emptied when you sign out | until you clear the site data in your browser | essential to the service you ask for |
assistant:recent:<id> | local storage | last five discussions used with the assistant (type, id, date) | until you clear the site data in your browser | interface preference |
tm.googleOauth | session storage | state of the Google sign-in in progress | for the life of the tab | essential to the service you ask for |
tm.pendingJoin | session storage | the code of a path to join, kept while you sign in | for the life of the tab | essential to the service you ask for |
tm.rootReloadOnce, pwa-install-dismissed | session storage | interface flags (one-time reload, install banner) | for the life of the tab | essential to the service you ask for |
tam.placement.v1 | session storage | progress and answers of the level test, without identity | for the life of the tab | essential to the service you ask for |
workbox-precache (Cache Storage) + service worker | browser cache | application files, for offline use | until the next update of the application | essential to the service you ask for |
transformers-cache (Cache Storage) | browser cache | speech-recognition model downloaded on first use of the microphone | until you clear the site data in your browser | essential to the service you ask for |
tam_audience_optout | cookie | your refusal of the audience measurement (see “Audience measurement”) | 400 days, renewed each time you open the refusal page | keeps your refusal |
cf_clearance, __cf_bm, cf_chl_rc_* | cookie | Cloudflare bot protection | 30 minutes of inactivity | set by Cloudflare for the security and reliability of its network |
NEL (Network Error Logging) | policy kept by the browser | Cloudflare asks your browser to report errors connecting to the Site | 7 days | set by Cloudflare for the security and reliability of its network |
The offline copies of your data stay on the device after a plain sign-out and are only erased when the account is deleted or when you clear the site data in your browser. Deleting the account does not erase the teacher onboarding, the list of the last discussions used with the assistant or the record that the device opened a learner’s link. On a shared device, clear the site data after use.
Security
All traffic is encrypted (HTTPS). The app never asks you for a password: sign-in relies on a single-use code, your Google account or a phishing-resistant passkey. Codes and refresh tokens are stored hashed; access tokens expire after 15 minutes and refresh tokens are single-use. Every change to a path is checked server-side: only its author can edit it. Cloudflare rate-limits requests per IP address and protects the Site against bots.
Minors
Tutors must be adults. Students may be minors. A student joins a path with the code or link their tutor shares, or their tutor creates an account for them without an e-mail address. By accepting the terms of service, a tutor undertakes to hold the authorisation of the holder of parental responsibility for any student under 15 they enrol, and to produce it on request; the app does not ask them to confirm it, and we do not check that undertaking. We collect no age data and ask students only for what is needed to follow their exercises.
In student mode, the audience measurement receives only the type of the pages viewed, with scrolling, time, clicks and loading speed, and no context (see “Audience measurement”); a student account counts in the service-usage statistics only in the number of accounts created each week, and a learner account created by a tutor never does. A parent or legal guardian can write to contact@teacher-and-me.com at any time to access, correct or erase their child’s data; they can also refuse the audience measurement in the browser their child uses and leave their child’s account out of the service-usage statistics.
Your rights
You have the right to access, rectify and erase your data, to restrict or object to its processing, to data portability, and to give instructions about your data after your death. These rights are set out in Articles 15 to 21 of the GDPR and, for instructions after death, in Article 85 of the French Data Protection Act (in French). Where processing relies on your consent, as keeping the channel code does, you can withdraw it at any time. You can object at any time to the audience measurement and the service-usage statistics, which rely on our legitimate interest.
From the application
- Delete my account (Settings): immediate deletion of your account (with the items used for the service-usage statistics), your paths (including those published in the catalogue, with their reviews), your enrolments, your progress and goals, your messages, your assistant discussions, your passkeys and your sessions, and the learner accounts you created; your catalogue comments are anonymised and hidden; the offline copies of your data are erased from the device used.
- Download my data (Settings): a JSON file with your profile, your paths (ids, titles, dates) or your progress and enrolments, your messages and the items used for the service-usage statistics. This export does not yet include the content of your exercises, your assistant discussions or your catalogue activity: write to us for a complete export. Tutors also have per-student CSV exports.
- Refuse audience measurement in this browser (Settings, under “Usage statistics”, or the button in “Audience measurement”) and Leave my account out of usage statistics (same place).
- Remove my answer to “How did you hear about Teacher and Me?” and Remove the channel code (Settings, under “How you found us”).
- Edit my name, hide my profile, erase my progress, delete my messages, clear an assistant discussion: in Settings and the relevant screens.
What account deletion does not cover yet
- the feedback you sent us (it contains your e-mail address): erased automatically after 14 days;
- the totals already computed by the audience measurement and the service-usage statistics, which hold no identifier and no account data;
- on the device used, the teacher onboarding (with your three answers to the welcome questionnaire), the list of your last discussions used with the assistant and the record that the device opened a learner’s link: clear the site data in your browser to remove them (see “What the Site stores in your browser”);
- the technical restore points of our databases, kept 30 days, then expired.
By e-mail
For any other request, write to contact@teacher-and-me.com from your account’s e-mail address (or with proof of identity). We answer within one month. If you believe your rights are not respected, you can lodge a complaint with the French supervisory authority, the CNIL, or with the authority of your country of residence.
Changes to this policy
The version in force is the one published at this address; its date and version number are at the top of the page. Every change to the Site that alters how your data is processed is reflected here on the same date. Only the French version is binding.
Version history
- Version 1.6 of 18 October 2026: the policy describes the audience measurement of the Site and the app and the service-usage statistics, which have run since 11 October 2026 without the policy describing them, the channel code, the question “How did you hear about Teacher and Me?” and the learner accounts a tutor creates; our PostHog project is deleted; the “Minors” section points to the undertaking tutors give in the terms of service; the table of what the Site stores in your browser is completed.
- Version 1.5 of 15 October 2026: tutors can no longer invite a student by e-mail address, and the pending invitations have been deleted from our databases with the addresses they held; the copy that earlier versions of the app kept in a tutor’s browser is deleted the next time the app opens.
- Version 1.4 of 15 October 2026: joining a path with the code or link its tutor shared enrols you on it like an invitation; that tutor sees your name, e-mail address and activity on that path.
- Version 1.3 of 15 October 2026: deleting your account now also erases your assistant discussions and usage counters.
- Version 1.2 of 11 October 2026: PostHog removed; the welcome-questionnaire answers no longer leave your browser; the location of the databases is stated more precisely (EU or UK). This version also said the Site no longer measured its audience, which was wrong: the audience measurement version 1.6 describes started the same day.
- Version 1.1 of 9 October 2026: the assistant discussions are now stored in the European Union.
- Version 1.0 of 23 September 2026: first version.