Privacy policy

Alphamingo is a small independent project — an app for learning English from the first letter, at any age: picture books, letters and their sounds, and a speaking helper — plus this website and a private admin tool used to write the books.

The data controller is Alphamingo, a small independent project, reachable at support@alphamingo.com. That address is read by the people who build the app, not by a support desk.

Many of the people using this app are young children, so this section matters most — and it applies to everyone, whatever their age.

We do not knowingly collect personal information from children, and the app is built so that there is nothing to collect. This is intended to satisfy the United States Children’s Online Privacy Protection Act (COPPA), Article 8 of the UK and EU GDPR, and the UK Age Appropriate Design Code.

Concretely, in the app there is:

• No sign-up, sign-in, or user profile. The app opens directly onto the books.
• No field anywhere that asks for a name, an age, a birthday, a photograph, or an email address.
• No advertising, no ad networks, and no behavioural or contextual ad targeting.
• No third-party analytics SDK, no crash-reporting SDK, no attribution SDK, and no social-media SDK. The app does send anonymous counts of which stories are read to our own database — section 3 says exactly what they contain.
• No profiling, no personalisation, and no recommendations based on a child’s behaviour.
• No chat, no comments, no sharing, no user-generated content, and no way for a child to be contacted by anyone.
• No links out of the reading experience, so a child cannot tap their way into a browser, a store, or a social network.
• No access to the camera, photo library, contacts, or location. The app never asks for these permissions, because it has no use for them.
• The microphone only for Say it in English and My voice, only while the button is held or recording is switched on, and only after your phone has asked your permission. What is said is not kept (section 3).
• No copy of anyone’s voice unless a grown-up chooses to make one of their own voice, after agreeing to it on the My voice screen. That screen says it is for adults, and it is not meant for children’s voices.
• No push notifications.

Because we hold no personal information about any child, there is no profile to sell, no profile to leak, and nothing for a parent to request or delete. If you believe a child has somehow provided us with personal information — for example by writing to our support address — email support@alphamingo.com and we will delete it.

Nothing that identifies you

No personal data that identifies you is collected by Alphamingo. The app stores your progress and the phrases you keep on your own device; they are never uploaded as a list, and deleting the app removes them.

The app also keeps the stories, pictures and recordings themselves on the device once it has fetched them, so they work with no internet. That is our content sitting in the app’s own folder — nothing about you — and deleting the app removes all of it.

Anonymous counts of what is read

So we can tell which stories readers enjoy and where they lose interest, the app sends a small count when the shelf of letters is opened, when a story is opened, when a page is reached, when the “Read to me” voice is used, and when a story is finished. It also sends one count the very first time it runs on a device, so we know a new installation has been opened; a reinstall simply counts again, as a new reader. A count contains only these things: which of those six events it was, the story’s short name (for example letter-d-duke), the page number, whether the device runs iOS or Android, the app’s version number, the reader id described next, and the time our database received it.

The reader id needs explaining, because it is the one identifier in this data. It is a random number the app invents the first time it runs and keeps on your device. It is derived from nothing — not the device’s hardware, not an advertising identifier, not a phone number, not an account — it is a coin flip your phone happens to remember. Deleting the app deletes it, and a reinstall invents a fresh one, so an id can be lost but never followed. We use it for a single calculation: how many different devices read a book, so we can tell “three readers finished this story” from “one very keen reader finished it three times”. It does not count people — a tablet shared by two people is one reader to us, and one person reading on both a phone and a tablet is two.

We keep the reader id because of one question the app stores cannot answer for us: does the same copy come back? The stores can tell us how many devices installed the app. Only the id can tell us that a copy which read a book on Monday opened another on Friday — and whether readers come back is the honest measure of whether the books are any good. That is all it is used for: counting devices, and counting returns. An earlier version of this policy called the id temporary; this is the change of mind, stated rather than buried.

The counts also carry the language your device is set to — “uk”, “pl”, “en” — as a two-letter code, with no country or region. We want to know whether Alphamingo is being used in homes where English is not the first language, because that is who it was made for. It is the device’s language, not yours, and it goes into the same totals as everything else.

Since 3 November 2026 each count is also marked with the country it was sent from, as a two-letter code such as “UA” or “ES”. Our database provider’s network works that out from the connection, the same way it does for this website; only the country is kept, never the IP address, the city or anything closer. It tells us which countries the app is used in — a market, not a location.

On Android, the first count — the one sent when a new installation is opened — also says where the installation came from, as Google Play reports it to the app: from a link on this website (and which one, for example “web:site” for the buttons on these pages), from Google Play itself, or from another link. It is the name of a link, not anything about the person who tapped it, and it is how we can tell whether this website brings anyone to the app. iPhones send nothing of the kind; Apple does not tell an app where it came from.

Beyond that id, a count contains no identifier — no device identifier, no advertising identifier, no account, and no location. So while we can now tell that the same installed copy of the app reached page 6 on Monday and finished the story on Tuesday, we still do not know whose copy it is, who was holding it, or where it was — and the id connects to nothing outside our own counts.

These counts go to our own database in Ireland. No third-party analytics company is involved, and there is no analytics or advertising SDK in the app. Under the UK and EU GDPR our lawful basis is legitimate interests — understanding which stories work so we can make better ones. Because the counts identify nobody — we could not pick out which rows came from your device even if you asked us to — there is nothing here to request or delete; if you would still rather the app did not send them, email support@alphamingo.com and we will tell you how to turn them off in the next release.

When you use Say it in English

You hold a button and say something short, usually in your own language. Three things happen, and none of them is stored by us:

• Turning speech into text. By default your phone does this with its own speech recognition, which Apple or Google may process on their servers under their own privacy policies; your phone asks your permission the first time. If you choose “Listen with Cloudflare” in the settings, the recording is sent to our server instead, turned into text by Cloudflare’s speech service, and deleted at once — it is never saved, logged or used for anything else.
• Making the English. The text of what you said, your chosen level, and the English of the phrases you have kept are sent to our server, which passes them to an AI translation service that works for us and writes the short English phrases. The service handles the text only to answer that one request and keeps nothing afterwards. The text is not used to train anything by us.
• What we learn from. For each phrase we keep a short note without your words: the languages, your level, the topic and area it was filed under (for example “Health · Hospital”), its length, which service answered and how fast, and the app’s random reader id. We also note when you star, replay or remove a phrase. This shows us which topics people need and what helps them learn.
• Only if you switch it on: your sentences. In Settings, “Help us improve” is off unless a grown-up (18 or over) switches it on. When it is on, we also keep the sentence you said and the answer, with the reader id, to make the answers better. Please don’t say names, addresses or anything private. Switch it off at any time; “Forget what I shared” on the same page deletes the sentences already kept.
• Saying it aloud. The English phrase is sent to Inworld, our voice service, which reads it in the voice you picked — one of the app’s characters or your own copy — and sends back the audio.

Under the UK and EU GDPR our lawful basis is that you asked for it — the feature cannot answer without the text — and our legitimate interest in running it. Under the US children’s privacy rule, a recording used only in place of typed words and deleted straight away is allowed without collecting anything about the child.

A copy of your own voice (adults only)

On the My voice screen a grown-up can read a short text aloud, about 10 to 30 seconds, so the English is spoken in their own voice. That recording is sent once, through our server, to Inworld, which makes the voice copy and keeps it. Neither we nor the app keep the recording itself once the copy is made. Only your phone holds the copy’s name, and it is used for one thing: speaking English back to you in the app. We never use a voice to identify anyone, and nobody else can use yours.

Our lawful basis is your consent, which you give by ticking “This is my own voice, and I am an adult” before recording. You can withdraw it at any time with Forget my voice, which deletes the copy at Inworld and on the phone. If you delete the app without pressing it first, email support@alphamingo.com with roughly when you made the copy and we will delete it.

Server request logs

Stories, illustrations and narration audio are downloaded from our hosting provider’s servers. Making that request unavoidably reveals your device’s IP address and general technical details such as the time of the request and the type of device — exactly as it would for any website or app that loads anything from the internet.

Our provider keeps these standard operational logs to run and secure the service. We do not use them to build a profile, we do not link them to any person, we do not use them for advertising or analytics, and we do not attempt to identify anyone from them. Under the UK and EU GDPR our lawful basis for this is legitimate interests — namely keeping the service running and defending it from abuse.

If you email us

We receive your email address and whatever you choose to write. We use it to answer you and for nothing else. Under the UK and EU GDPR our lawful basis is legitimate interests — replying to someone who has written to us. Please do not include a child’s personal details — we do not need them to help you.

If you send us feedback from the app

Settings has a Leave us feedback page. If you use it, we receive what you write, the language you use the app in and the one you are learning, whether it is an iPhone or Android and which version, and the app’s random reader id described above. We use it to read your message and answer it in the app, and for nothing else. Only the phone that sent a message can read our answer. It is not linked to your name or email, because the app never asks for them. Under the UK and EU GDPR our lawful basis is legitimate interests — replying to someone who has written to us. Please do not include a child’s personal details.

From the administrator

The adults who write the books have accounts, each with an email address and a password. See section 6.

Plainly, and without exceptions hidden in other sections:

• We do not sell personal information. We never have and there is nothing to sell.
• We do not share or disclose personal information to third parties for their own purposes.
• We do not show advertising of any kind, and we do not work with advertising networks or data brokers.
• We do not use third-party analytics in the app. The counts described in section 3 go to our own database, carry no identifier beyond the random reader id described there, and are never shared.
• We do not keep what you say to Say it in English, and we never use anyone’s voice to identify them.
• We do not track users across other apps or websites.
• We do not use cookies or similar technologies in the app.
• We do not build advertising profiles, and we do not make automated decisions that have any effect on anyone.

Illustrations are original drawings. Most are simple flat shapes stored inside the story itself; painted books are ordinary picture files the app downloads from our own storage, the same one-way download as the audio below. There are no photographs of real people anywhere in the app.

The “Read to me” button works one of two ways:

• A recorded audio file. Where a book has narration, the app downloads an ordinary audio file from our storage and plays it. This is a one-way download; nothing about you is sent with the request beyond what section 3 describes. These recordings are made with an AI voice, not by a human reader. Nothing a reader says or does is sent anywhere to produce them — they are made once, by us, before the book is published.
• Your device’s own built-in voice. Where a page has no recorded narration yet, the app asks your phone or tablet’s operating system to speak the sentence, using the same speech feature the device already provides. We hand the sentence to the device; we do not send it to us. How your operating system performs that speech is governed by Apple’s or Google’s own privacy policy.

The books never listen. Reading a book only ever plays sound. The microphone is used by Say it in English and My voice alone, as section 3 describes.

The books are written and edited through a private page on this website. Only the adults on our allowlist can use it — today there are two. Each account holds an email address, a password stored in hashed form by our authentication provider, and the times the account signed in.

There is no public sign-up. A visitor who finds the admin page sees a sign-in form and cannot create an account. People using the app have no account, never see this page, and are not affected by it.

We keep the list of outside companies as short as we can. Today it is:

• Supabase — stores the books, the narration audio files, and the single administrator account. Supabase acts as our data processor and does not use any of it for its own purposes.
• Our website host — serves this website and the admin page, and keeps standard access logs.
• AI translation services — write the English for Say it in English. They work on our behalf, receive only the text described in section 3, use it only to answer that one request and keep nothing afterwards. We may use more than one, so the app keeps working when one is busy.
• Cloudflare — if you choose it, turns a recording into text for Say it in English. It processes the recording for that one request and keeps nothing for itself.
• Inworld — our voice service. It speaks the English phrases and makes and stores a voice copy when a grown-up chooses to make one.
• Apple or Google — your phone’s own speech recognition, when the app listens with the phone. Their privacy policies apply to it.
• Our email provider — carries messages you send to our support address.
• Google (Google Analytics) — counts visits to this website, as described in section 10. It is not used in the app.

To be specific about where things sit: the books, the narration audio and the administrator account live in a Supabase project hosted in the EU (Ireland). This website and the admin page run on Cloudflare Workers, which serves requests from whichever of its locations is nearest the visitor, so a request made outside Europe is handled outside Europe. The AI translation services may handle the text of what you say in the United States. Where personal data is transferred internationally we rely on the safeguards those providers put in place, such as standard contractual clauses. Write to support@alphamingo.com if you would like the current specifics.

We will also disclose information if the law genuinely requires it — for example a valid court order — and we will resist requests that go further than the law requires.

• Data that identifies people using the app: none is asked for. A sentence shared through Help us improve could say who you are if you put that in it — please don’t; Forget what I shared deletes it.
• What you say to Say it in English: not stored, unless a grown-up switched on Help us improve. Recordings are never stored. The notes without your words (section 3) are kept for up to 12 months.
• Sentences shared through Help us improve: kept for up to 12 months, then deleted automatically, or at once when you press Forget what I shared.
• A copy of your own voice: kept by Inworld until you press Forget my voice, or until you ask us to delete it.
• Anonymous counts of what is read: kept for up to 18 months and then deleted automatically, so we can compare one season with the next. They identify nobody — the reader id inside them is a random number that leads nowhere (section 3) — so this is a housekeeping limit rather than a privacy one.
• Counts of visits to this website: deleted automatically after 90 days. The visitor code inside them stops meaning anything at the end of the day it was made — section 10 explains it — and 90 days of counts just lets us compare one month with the one before.
• Server request logs, which contain your device’s IP address: kept so we can keep the service running and safe, and deleted automatically by our providers — Cloudflare keeps them at most 7 days, Supabase 1 day on our current plan. We do not copy them anywhere else, so nothing survives that window on our side.
• Administrator accounts: kept while that person is still writing books, so they can sign in, and deleted on request.
• Support emails: kept while we are helping you and for up to 12 months afterwards, in case you write again about the same thing, then deleted.
• Feedback sent from the app: kept while we answer it and for up to 12 months afterwards, the same as support emails, then deleted. Email us and we delete it sooner.

Depending on where you live — including under the UK and EU GDPR, and under California law — you have the right to ask what personal data is held about you, to get a copy, to have it corrected or deleted, to restrict how it is used while a question about it is settled, to object to how it is used, and to complain to a regulator.

In practice, the honest answer for almost everyone reading this is that we hold nothing about you, so there is nothing to retrieve or erase. The exceptions are a voice copy you chose to make, which Forget my voice deletes, and any feedback you sent from the app, which we delete when you ask. If you would like to be certain, email support@alphamingo.com and we will tell you exactly what we do and do not hold.

To remove everything on your side, press Forget my voice if you made one, then delete the app. That removes the books, your kept phrases and any progress stored on your device. Nothing else about you remains on our servers, because nothing else was ever put there.

If you are in the UK or EEA and think we have handled data badly, you may complain to your national data protection authority. We would rather you told us first at support@alphamingo.com so we can fix it.

The public pages of this website — this one, the home page and the support page — carry no advertising and no tracking pixels. Since 3 November 2026 they use Google Analytics to count visits, as described below. Apart from that, the only cookies this site sets are the ones strictly necessary to keep the administrator signed in to the private admin page, and you will never receive one unless you sign in there.

Google Analytics

To understand which countries our visitors come from, how they find us and what they read, the website uses Google Analytics, a service of Google. It sets cookies in your browser so it can tell a first visit from a return visit, and it records the pages you view, how long you stay, the website or search that sent you, your device type and browser, your approximate location (country and city, worked out by Google from your IP address), and taps on our App Store and Google Play buttons. We switched off Google signals and ad personalisation: this is counting, not advertising. We see only totals, never a person.

If your device is set to a European time zone, the site asks you first and Google Analytics does not load at all unless you press “Yes, count it”; “No thanks” is remembered on your device and nothing is sent. Elsewhere it loads with the page. Under the UK and EU GDPR and the ePrivacy rules our lawful basis is your consent; elsewhere it is our legitimate interest in understanding how the site is used. You can change your mind by clearing this site’s data in your browser, which brings the question back, or by using Google’s browser add-on that turns Google Analytics off. Google keeps the detailed records for two months; the totals it produces from them are kept longer. Adults signed in to the admin page are never counted.

The website does count its visitors, and this is the plain description of how. When our server hands you a page, it records five things: which page it was — the path only, never the query string, which is the part of an address where anything personal could hide; a visitor code; the website that sent you here, as a bare host name such as “google.com” and never the page or the search you typed; the two-letter country the request came from, which is a market and not a location; and whether the request came from a phone, a tablet or a computer — one of those three words, never the model or the browser. The visitor code is a one-way scramble (a SHA-256 hash) of a secret only our server knows, the current date, your IP address, and the technical name your browser sends with every request. Only the scrambled result is stored; your IP address and browser name themselves never are.

Because the date is inside the scramble, the same visitor produces a completely different code tomorrow, and the codes cannot be joined across days: they can answer “how many different visitors came today” and cannot answer “did this person come back”. There is no cookie and no other company involved in this counting — it happens on our own server while it serves the page, separately from Google Analytics. These counts are deleted automatically after 90 days.

One thing is kept on your own device, and only there. When you open a story on this website, your browser writes down which book it was, so the home page can offer you “Carry on reading”. It is a short list of book names — at most eight — held in your browser’s own storage on your device. It is never sent to us, it is not part of the counting described above, and it is not a cookie: nothing about it travels with your requests. Clearing this site’s data in your browser deletes it, and if your browser blocks that storage the shelf simply does not appear.

If this policy changes, the date at the top changes with it. If a change ever meaningfully affects what we collect — which we do not intend — we will say so clearly on this page rather than quietly editing a sentence.

Two such changes have happened, and both are spelled out above rather than hidden: on 6 August 2026 the app began sending the anonymous reading counts described in section 3, and on 17 August 2026 those counts gained the reader id described in section 3, and this website began counting its visitors as described in section 10. On 2 September 2026 the reader id stopped being called temporary, and the counts gained the device’s language — both in section 3.

On 1 November 2026 two more changes were made, both in sections 2 and 3: Alphamingo is now described as an app for learning English at any age rather than only for children, and the app gained Say it in English and My voice, which use the microphone. The earlier sentence that the app never asks for the microphone was removed because it is no longer true. The same day, on Android, the first count began saying which link the app was installed from, as described in section 3.

On 3 November 2026 the app’s anonymous counts began to carry the country they were sent from, as described in section 3. The same day the app gained a Leave us feedback page in Settings, described in section 3, with how long feedback is kept in section 8. The same day it began teaching German and Spanish as well as English; that changes nothing about what is collected. Also on 3 November 2026 the website began using Google Analytics, with a question first for visitors in Europe, as described in section 10; the earlier sentence that the website has no third-party analytics and no cookie banner was removed because it is no longer true.

On 4 November 2026 sections 3 and 7 began describing the AI service that writes the English as one of several AI translation services working for us, possibly in the United States, so the app can keep answering when one of them is busy. What is sent stays the same. The same day the app began keeping a note of each phrase without your words — languages, level, topic, area — and, only if a grown-up switches on Help us improve in Settings, the sentence and its answer, as described in sections 3 and 8. The earlier sentence that what you say is never stored was changed because of that switch.

Questions, corrections, or anything at all about this policy: email support@alphamingo.com. A real person reads it.

Last updated: 4 November 2026.