JavaScript is disabled, so this page stays in English and follows your device theme.

Privacy policy

Last updated: 11 October 2026

Our commitment to your privacy

Matn is built with privacy as a core principle. Your study is personal, and we have no interest in tracking, profiling, or selling your data.

What we don't collect

  • No third-party advertising analytics or user profiling (this website uses no tracking scripts)
  • No cookies (this website doesn't use any)
  • No third-party advertising networks
  • No user accounts or login systems (on this website)

Waitlist

If you join the beta waitlist, we collect your email address, platform preference, language, submission timestamp and consent record, and, if the link you arrived by carried a source tag (such as ?ref=ig), that tag. Your address is not confirmed until you follow the confirmation link in our email. The record is stored in Cloudflare D1. Cloudflare processes the request IP through Rate Limiting to prevent abuse, but Matn does not store the IP address.

  • We use confirmed addresses to confirm a waitlist entry and may send occasional Matn updates. We do not promise beta invitations or platform-availability notices by email
  • Cloudflare processes the request IP through Rate Limiting to prevent spam and abuse; Matn does not write the IP address to the waitlist record. Resend receives the email address only to send waitlist emails and reports bounce, complaint, and suppression events so we can avoid sending further email to unavailable or complaining addresses
  • A confirmation link expires after seven days. Unconfirmed signups are deleted after that. Consent for a confirmed signup lasts until you withdraw it, with no separate expiry, and the unsubscribe link in every email keeps working for as long as you keep it. When you unsubscribe, we delete your address from the waitlist immediately and keep only a one-way hash of it, so it can never be mailed again
  • Every waitlist email includes an unsubscribe link. You can also request a copy of, or deletion of, your waitlist record at any time by emailing hello@matn.dev from the same email address used to join

Text and feature requests

If you use the suggestion form, we collect the request category, message, optional email address, language, and submission time. The record is stored in Cloudflare D1. Cloudflare processes the request IP through Rate Limiting for abuse control, but Matn does not store it. We use an optional email address only if a reply about that request is useful; submitting a request does not add you to the mailing list.

How the app works

The Matn app keeps core study data on your device so you can read and review offline:

  • Texts and available reference audio can be used offline. Recitation tracking requires a one-time download of the Arabic recognition model to your device
  • Your progress, review history, and settings are stored locally using on-device storage
  • Matn uses your microphone to record your recitation. Saved recordings stay on your device unless you are signed in and Cloud sync is on, in which case Matn uploads them to Firebase Storage for your other signed-in devices.
  • On-device recitation recognition follows the words Matn recognises during practice. It does not assess pronunciation or replace a teacher. Your voice audio is processed on your device by default. You may optionally choose to share recordings to help improve speech recognition (see "Data You May Choose to Share" below)
  • Optional sign-in and Cloud sync use Firebase Authentication, Cloud Firestore, and Firebase Storage. With Cloud sync on, Matn syncs selected study data and saved recordings across your signed-in devices.
  • Sign-in is optional. If you choose to sign in, you may use Sign in with Apple, Google Sign-In, or an email and password. Apple and Google act as identity providers and pass us your email address and, where you allow it, your name. We do not receive your password or your Apple or Google account credentials. Apple offers a private email relay if you prefer not to share your real address
  • Usage analytics (such as which features are popular) are collected via Firebase Analytics only if you opt in under "Usage data and crash reports" in Settings. Analytics is off by default and can be turned off at any time in the same place. If you sign in, analytics may be associated with an opaque Firebase user ID, but never with your email, name, text content, or recordings. When analytics is on, Google Analytics infers an approximate location, such as your country or city, from your device’s network address. Matn does not collect your precise location
  • Crash reports and diagnostics may be sent to Firebase Crashlytics on supported Apple platforms and Android only if you opt in under "Usage data and crash reports" in Settings. Crash reporting is off by default and shares the same switch in Settings. If you sign in, reports may include an opaque Firebase user ID, but not your email, name, text content, or recordings
  • The app uses random identifiers that are not tied to your name or email address. If you report a problem while signed out, a random installation identifier kept on your device is sent with the report so that we can limit abuse. If you opt in under "Usage data and crash reports", Firebase Analytics and Crashlytics also use the Firebase installation identifier of your copy of the app. Matn does not use your device’s advertising identifier

Data you may choose to share

Some features let you optionally share data. Each requires separate, explicit consent:

  • Optional feedback: When you submit feedback, you may choose to attach a screenshot to help us investigate and to add an email address. Both are optional. They are processed with your feedback text and basic diagnostic details
  • Recording contributions: You may opt in via the "Contribute recitation clips" setting to share clips that help evaluate speech recognition. Clips are used to train Matn's recitation recognition model only if you also turn on "Use my clips for training". That covers clips you share after turning it on, never earlier ones. Each clip carries the recognised text, up to 40 word samples, and device metadata alongside the audio. Contributed recordings are account-linked personal data, not anonymous data, and are deleted automatically after 18 months. Turning this setting off stops new uploads; it does not delete clips or metrics already shared. To remove them, request deletion in the app. This starts an asynchronous deletion job, and removal is complete only when the job finishes
  • Recitation metrics: When you opt in to "Contribute recitation clips", Matn may also record the selected text's identifier, line and word positions, and limited session metrics such as reveal, silence, no-speech, and line-advance events. These metrics do not include audio, transcripts, recognised text, or canonical text. They are account-linked personal data. Turning the setting off stops new uploads; it does not delete metrics already shared. Request deletion in the app to start an asynchronous deletion job; removal is complete only when the job finishes
  • Issue reports: If you report a problem with a bayt, we collect the bayt position, error type, and an optional note. For a recitation-tracking report, this includes any correction you make. The expected words and the words Matn recognised are included only if you have turned on "Use my clips for training", and are then used to train speech recognition. If you are signed in, reports are linked to your account and kept until you delete your account. If you are signed out, you may add a contact email, and the report is kept for up to 18 months; we also keep a short-lived hash of your device's network address, for up to 7 days, to limit abuse

Training use

If you turn on "Use my clips for training", Matn may use the clips you share from then on, with their transcripts and word samples, and the expected and recognised text in recitation-tracking reports you file, to train the speech recognition model that ships in the app. This is separate from sharing clips for evaluation and is off unless you turn it on. You can turn it off at any time in the app's Privacy settings; material shared afterwards is not used for training. Deleting your contributed data removes it from future training, but a model already trained on it is not retrained.

Voice recordings

Recordings stored only on your device stay there. Recordings uploaded through Cloud sync or "Contribute recitation clips" are account-linked personal data, not anonymous data. Matn does not use your voice to identify you, sell recordings, or use them for advertising.

Contributor recording portal

Reference recordings can also be contributed through a separate recording portal, not the app itself. If you submit a recording there:

Recordings made on the recording portal are covered by the portal's own agreement: by contributing, a reciter agrees that Matn may play the recording to learners, credit it as they choose, and use it to train and evaluate Matn's recognition. This differs from the app, where training on practice clips is a separate choice.

Without an account, the portal's help page (matn.dev/help) takes single-bayt clips. These clips are used only to train and test Matn's recognition of recitation; they are never played to learners or published. No name, email or account is attached: an anonymous identifier is kept in your browser, and from that browser you can delete any or all of your clips at any time, though a model already trained on a clip is not untrained by its deletion. Clips are kept for 18 months, then deleted. Each browser can send up to 60 clips a day.

  • If your recording is approved after review, it is published in Matn with the credit you chose on the consent form: under your name or kunya, or anonymously
  • An approved recording is kept because it is part of the published corpus. A recording that stays unreviewed is deleted automatically 180 days after upload, and a rejected recording is deleted automatically 90 days after rejection
  • Every font on the portal page is self-hosted; none are loaded from an external service
  • The portal uses Firebase App Check (reCAPTCHA v3) to stop abusive automated uploads, which may load its own script from Google when verifying you

Data retention

  • Personal recordings (cross-device sync): retained while your account exists. An account-deletion request starts removal of cloud data linked to that account; removal is complete only when the deletion job finishes
  • Contributed recordings: deleted automatically 18 months after contribution. Withdrawing consent stops new uploads; it does not delete clips or metrics already shared. A deletion request in the app starts an asynchronous job that deletes them sooner, and removal is complete only when the job finishes
  • Usage analytics: collected only after opt-in and retained under the standard Firebase Analytics retention period configured for our project
  • Crash reports: collected only after opt-in and retained under standard Firebase Crashlytics retention (90 days)

Service providers

We do not sell your data and we do not share it for advertising. To run the optional cloud features, the waitlist, and crash diagnostics, we rely on the following providers, who process data on our behalf under their own terms:

  • Google (Firebase): provides Authentication, Cloud Firestore, Storage, Crashlytics, Analytics, and App Check. Firebase stores your synced data, contributed recordings, issue reports, crash reports, and usage analytics
  • Apple: provides Sign in with Apple as an identity option. Apple passes us your email (or a private relay address) and, where you allow it, your name
  • Cloudflare: hosts this website, stores the waitlist and text or feature requests in D1, delivers waitlist email work through Queues, processes request IP addresses through Rate Limiting for abuse control, and delivers the recognition model the app downloads
  • Resend: delivers waitlist emails and reports bounce, complaint, and suppression events so we can avoid further email to unavailable or complaining addresses
  • Wiredash: provides optional in-app feedback. Wiredash receives nothing until you open the feedback flow. Opening it sends Wiredash a random identifier, your operating system and its version, your language, and the app version; submitting feedback sends your message and basic diagnostic details such as device model and screen size, with a screenshot or an email address only if you choose to add them

Lawful basis for processing

Where the GDPR or similar laws apply, we rely on these lawful bases:

  • Account sign-in and sync: performance of a contract, to provide the service you asked for
  • Waitlist emails, usage analytics, crash reporting, and contributed recitation clips and their associated metrics: your consent, which you can withdraw at any time
  • Training use of contributed clips and recognition reports: your consent, which you can withdraw at any time
  • Signed-out feedback and issue reports, including an optional contact email: our legitimate interest in investigating and fixing the problem you reported
  • App Check attestation (Play Integrity, App Attest, or the recording portal's reCAPTCHA v3): our legitimate interest in preventing abuse of our backend services
  • Waitlist security processing, including Cloudflare's processing of the request IP: our legitimate interest in preventing spam, abuse, duplicate submissions, and misuse of the waitlist

International data transfers

Our providers (Google Firebase, Cloudflare, Resend, and Wiredash) operate globally, so your data may be processed on servers outside your country, including in the United States. Where required, these transfers rely on the safeguards offered by each provider, such as the European Commission's Standard Contractual Clauses. Data is encrypted in transit: the app and this website send it to these providers only over HTTPS (TLS).

Children

Matn is intended for a general audience and is not directed at children under 13 (or the minimum age of digital consent in your country, where that is higher). The app does not include an age gate, so we ask that children use it only with the involvement of a parent or guardian. If you believe a child has provided us with personal data, contact us at hello@matn.dev and we will delete it.

Data controller

The data controller for Matn is the Matn project, reachable at hello@matn.dev. For any question about this policy or your data, please use that address.

Website privacy

This website (matn.dev) is a static site with minimal server-side processing. Your language and theme preferences are saved in your browser's local storage for convenience, but this data never leaves your device. If you arrive by a link with a source tag (such as ?ref=ig), the tag is kept in this tab’s session storage until the tab closes and is sent only with a waitlist signup. The waitlist and suggestion forms send data to a Cloudflare Worker as described above.

Your choices & rights

You stay in control of your data:

  • Offline use remains available even if you never sign in
  • Cloud-backed features are optional and depend on the choices you make in the app
  • You can withdraw consent for recording sharing at any time in the app's Privacy settings. This stops new uploads; it does not delete data already shared
  • You can turn off training use separately at any time in the same settings; turning off recording sharing also turns it off
  • You can request deletion of contributed data from within the app. This starts an asynchronous deletion job; removal is complete only when the job finishes. You can also delete your entire account to start removal of cloud data linked to that account
  • You can request a copy of your data or ask questions by emailing hello@matn.dev
  • If you believe your data rights have been violated, you may lodge a complaint with your local data protection authority

Contact

If you have questions about privacy, email us at hello@matn.dev