This is a translation for convenience. The Spanish version prevails. Versión en español

CoffeEZ Privacy Policy

Last updated: October 25, 2026

CoffeEZ is developed by an independent developer. We process the minimum data needed for the app to work. We do not sell personal data, we do not show ads and we do not build commercial profiles.

Data controller

Controller: Marcos Menjón Casanova
Contact: privacidad@coffeezapp.com

1. Use without an account

You can use CoffeEZ without signing up to save your coffees and recipes, which in that case are stored only on your device. The artificial intelligence features (scanning labels or technical data sheets, generating recipes and cupping adjustments) require an account.

If you have Android backup turned on, the system may include in your Google account’s backup, and when you move your data to a new phone, what the app stores on your device about your coffees, your recipes and your settings. It does not include the photos of your coffees, your session or the access keys to the services the app uses, your preferences about usage statistics and crash reports, or cached copies. When you restore it you will have to sign in again, the app will ask you again about usage statistics, and your decision about improving the recommendations (section 13) will be taken from your account, not from the backup. We have no access to that backup: you manage it from Android’s settings.

Even if you do not sign up, the first time you open the app on a device, and every time you sign out, an anonymous session is created in Firebase Authentication: a random identifier, with no email or name, with which the app reads the service configuration and the roasters’ official packages. If you create an account, that identifier becomes your account’s; if you sign in to one you already had, it is discarded. In any other case, it is deleted automatically after 90 days without using the app.

When you sign out, the app removes your account’s data from your device: your coffees, recipes and photos, your catalog if you are a roaster, the artificial intelligence usage counter, your decision about improving the recommendations (section 13) and cached copies. They remain stored in your account and come back when you sign in again. The device’s own settings stay, such as the theme, the sound or your grinders. The version history of each recipe is kept only on your device and is deleted when you sign out.

2. What data we process and why

Unless otherwise stated, the legal basis is performance of the contract (Art. 6.1.b GDPR): we need this data to give you the service you ask for.

• Account. Your email address and the user identifier that Firebase assigns, to identify your account, sync your coffees and recipes between devices and enable the artificial intelligence features. If you sign in with Google, also your name and your Google profile photo, which are only shown in Settings. If you sign up with email and password we do not ask for your name, and we send you a message to verify that the address is yours: until you verify it, you cannot use the artificial intelligence features or apply for the Partner Roaster account.
• Coffees and recipes. What you save in the pantry and in your recipes. With an account they are stored in our database; without an account, only on your device.
• Photos of your coffees. The photo you associate with a coffee in your pantry or, if you are a roaster, with a batch in your catalog. It can be one you choose from your gallery or the label image itself: when you scan a label with the camera or from the gallery, the app suggests that image as the coffee’s photo, and you can remove it before saving. Without an account, the photo stays only on your device. With an account, it is uploaded compressed to your private storage folder; only you can access it.
• Photos and PDFs of labels. When you scan a coffee’s label or technical data sheet, the image or the PDF is sent to the artificial intelligence to read its data (see section 3). Our servers pass them to Gemini and do not store them. The label image is also suggested to you as the coffee’s photo (previous item); a PDF is not. In earlier versions of the app a photo of the beans was also sent to estimate the roast, with the same treatment; the current version no longer asks for it.
• Roaster application. If you apply for the Partner Roaster account: your account’s email, the name of the roastery or company, its NIF/CIF or other tax identifier and the link to your website or your Instagram, with the date of the application (and, if you apply again after a decision, that of the first one and the attempt number). Your application is reviewed by hand by the data controller from a private panel on our website, where they also see your account data (your email, whether it is verified, how you sign in, when you created it and whether it is disabled) and, from your profile in the app, your plan and your roaster status. To decide they may consult public sources: your website, your Instagram profile, Google Maps and, if your identifier is from the European Union, the European Commission’s VIES service, to which they enter that identifier. While the application is pending, our servers also run these automatic checks: the first and the second, on every review (every 15 minutes); the third to the fifth, which consult external sources, at most once a month (earlier if you change the link, the identifier or the country, if some source did not answer or if the controller requests it from the panel). First: they deduce from the link whether it is a website of your own, an Instagram account or a page on another platform, and store that type, the domain or the account and a simplified form of the name (without accents, signs or legal form). Second: if your account is an email and password one, is verified and the domain of your email is that of the website you give, they note it as proof that you control that domain; a Google account does not work for this proof. Third: if your tax identifier is from the European Union, they check it in VIES and store whether it is recorded as valid and whether the name VIES returns matches the one in your application, without storing that name. Fourth: if the link is a website of your own, they check in the public DNS whether its domain has email (MX records) and, in the domain’s public registry (RDAP), its registration date; they store only those two data. Fifth: if the link is a website of your own or a page on another platform, they visit that address (following at most three redirects) and, if it is linked from there, a page of the same site whose address indicates it is a contact, about-us or legal page (such as the legal notice), like a normal visit to a public page, although they only download the text of each page, without images or scripts, with a maximum time and size and without storing its content: they note whether it responds and to which domain it leads, its title, whether it talks about coffee and mentions the roastery’s name, which Instagram account it links to, which was the second page and the email addresses that appear on those pages (at most 10), with the page and the place where they appear and whether they are from your website’s domain; we delete those addresses when the application is decided. A link to Instagram is not visited. In those queries, VIES receives only your tax identifier; the DNS and the domain registry, only the domain name; and the sites that serve those pages (your website or the platform, and those they redirect to) receive a visit from our servers, identified as CoffeEZ. The results of the first and the second are stored in your application; those of the third to the fifth, in the part that only the controller sees. To prove that the website or the Instagram account you give are yours, you can request from the app a six-digit verification code, which we send by email to an address that you choose and that must be on your website’s domain, if your link is a website of your own, or the one published by your Instagram profile or your page on another platform, if it is not. We can also send it to you ourselves from the panel, to an address of your business chosen by the controller: one on your website’s domain, one of those that appear on your website or in your profile (such as those the fifth check finds) or the one you tell us by email; it does not have to be on your website’s domain and there may be more than one per application. The email is sent from verificacion@coffeezapp.com and carries the code, your website or your Instagram account and your account’s email masked (for example, j•••@gmail.com), without your name or any other text; whoever receives it without having asked for anything can write to us so that we block the application. We store in your application the address each code went to (the full address only while the application is pending; once decided it is left masked), when it was sent, how many you have requested and how many confirmation attempts you have made, and the code only as a cryptographic summary (salted hash) with which we check the one you type, never the code itself. A code requested from the app is valid for 30 minutes and one sent from the panel, for 7 days. To prevent abuse we limit the codes you request from the app per account (3 per day and 10 in total, to a single address in each application: if you apply again after a decision, you can use another) and, if the code goes to an email on your website’s domain, per domain (2 accounts per day and only one with valid codes), the total sends per day (also those from the panel) and the confirmation attempts from the app (5 per day); those counters are stored in your application; the domain ones, in a record of that domain that carries neither your identifier nor your email, but a random mark whose relation to you appears only in your application, and which expires at the latest 2 days after the day you requested the code and is then deleted automatically, normally within the following 24 hours; and the total, in a daily counter with no data about you. If the code went to an email on your website’s domain, its confirmation is noted as proof that you control that channel; if it went to another address (a published one or the one you told us by email), the controller checks by hand that it appears on your website or in your profile before accepting it; if it does not appear, it does not count as proof. Whoever receives the code email without being a user (whoever reads that mailbox of your business) receives only what is described, and the code travels unencrypted inside the email and, if you reply to us by email, in your reply. The emails are sent by Mailjet on our behalf (section 5). In addition, every time the controller opens the panel, it cross-checks your application with the others, also after deciding it and for as long as you keep the account, to detect whether several accounts request verification for the same roastery, domain, Instagram account or tax identifier. If we need more information, we write to your account’s email; the question is noted in your application, and your reply, which reaches us by email, we copy by hand into it (the conversation keeps at most the last 50 messages). We also store the decision (approved, rejected or verification withdrawn) with its date and reason, the controller’s internal notes and a history of the actions on your application (at most the last 100), with the date and who did it. While your application is under review, the app shows, in the versions that have the “Your application” screen (in Settings), its status, our question if there is one and whether you have already proven that the website or the profile are yours (with the code, with the link in the bio or with your account on the website’s domain); in earlier ones, only that it is under review. We also send you the questions by email, and we communicate the rejection and the withdrawal of verification to you by email, with their reason. If you apply again, the previous conversation and decision are deleted from the application; in the history remain the decision with its reason and our questions, and of your replies only the date on which we recorded them. The notes and the history are not shown in the app; you can request them with your right of access (section 10). Legal basis: performance of the contract to process your application (Art. 6.1.b GDPR) and, for the checks against impersonation of roasters, our legitimate interest in having official packages published only by real roasters (Art. 6.1.f GDPR).
• Verified roaster seal. If we approve your application, we create your seal: the name under which you appear on your official packages, the date of approval, that of your last publication and, if we note it, the country of your roastery. We set the name when approving (normally that of your roastery, as you stated it) and we may correct it afterwards. Whoever opens one of your packages from the app reads your seal, also without an account: it is what verifies that the package is official. If we change the seal’s name, it also changes in the packages you already published. If we withdraw the verification, we delete the seal and your QR codes stop working; your packages are deleted or kept without a seal, and we tell you in the email. The seal and the packages are deleted if you delete your account (section 11).
• Roaster application: optional data and the link in the bio. If you submit the application from a version of the app with the new form, in addition to what is described in “Roaster application” you can state the roastery’s country (it decides whether the tax identifier is mandatory, as in Spain, Portugal and Italy, and in which country we look it up in VIES), its city, your website and your Instagram account separately and a message for us. Everything is optional except the website or the Instagram. They are stored in your application together with the mark that it came from the new form and its date, and the controller sees them in the panel; if you later submit the application again from an earlier version of the app, that data is deleted. If you only have Instagram and your profile does not publish an email, you can prove that the profile is yours with a link for your bio: from the app you request a random six-character code, valid for 14 days, and you put it in your bio as coffeezapp.com/?v=XXXXXX. We store in your application the code, the account you requested it for and their dates. The controller opens your public profile, checks that the link appears and confirms it in the panel by pasting the profile’s address; it is then noted as proof of control of that profile (the account, the code and the date) and the pending code is deleted, as when the application is decided or if you apply again. You can remove the link from your bio as soon as it is confirmed. Legal basis: the same as for the roaster application.
• Roaster catalog. If you are a Partner Roaster, the batches in your catalog, with their coffee and their recipes, and a record of the catalog’s activity (for example, batches created, edited or archived, labels scanned and packages published). Only you see them; what you publish in a package with a QR code can be opened by anyone who has the code.
• Openings of roasters’ QR codes. When someone opens the QR code of a roaster’s package, from the app or from the web, we add one opening to the counters of that package and of its roaster, at most once a day per device or browser. The counters do not store data about who opens it, and the roaster sees only the figures. So as not to count twice on the same day, the app stores on your device, and the website in your browser’s local storage (it is not a cookie), the package’s code and the date of the last scan; that date is not sent to our servers. Each opening also carries a random one-time number, which does not identify you, so as not to count the same one twice if the request arrives repeated; we store the latest ones of each day next to the package’s counters. To prevent abuse, each package counts at most a fixed number of openings per day. Legal basis: legitimate interest in measuring the Partner Roaster program (Art. 6.1.f GDPR).
• Arrival through a roaster’s QR. If you create your account after opening the QR code of a roaster’s package (or had created it at most 24 hours earlier), we store in your profile which roaster and which package it was and the date, and we add one new account to that roaster’s statistics. Until you create the account, the code of the first package you opened and the time stay only on your device. To the roaster we show only how many new accounts arrive, never who you are. If that roaster deletes their account, we remove from your profile which roaster and which package it was. Legal basis: legitimate interest in measuring the Partner Roaster program (Art. 6.1.f GDPR).
• Interest in the plans. While the paid plans are not on sale, if you tap “Notify me” on the plans screen and you have an account, we store in your profile which plan interests you, where you came to that screen from (for example, from Settings or when running out of artificial intelligence uses) and the date, to notify you when they are available and to know which plans are of interest. The notice is given only inside the app: we do not use your email for that.
• Purchases. If you buy Barista (one-time payment) or Catador (monthly or annual subscription), or make a donation, Google Play and RevenueCat tell us, associated with your user identifier, what you have active: whether you bought Barista, until when your Catador subscription is active and whether you are in the trial period and since when. We store it in your profile to apply your plan. Of the amount of each purchase we only store the sum of each month’s revenue, without linking it to your account. We have no access to your payment details or your card.
• Interactions with the app (Firebase Analytics). Usage events of these types: the use of artificial intelligence (label scans started, successful or with an error, the first scan on each device, recipes generated and the usage limit reached); general use (the introduction completed and the coffee limit reached); the plans (the plans screen shown, interest in a plan, purchases started, completed or cancelled and trials started); and sharing (recipes shared, batches published by a roaster and accounts that arrive from a roaster’s QR, without saying which one). The events carry only closed categories, such as the recipe’s method, the plan or the origin of the scan. In addition, Analytics’ automatic events (such as the first open, sessions and app updates), the device model, the Android and app versions, the language and the approximate region deduced from the connection. They are associated with a random installation identifier, not with your account, and do not include texts you write, coffee names or your email. They are only collected if you accept: the app asks you once and you can change it whenever you want in Settings → “Usage statistics”. Until you answer, and if you do not accept, nothing is collected. If you withdraw it, we stop collecting them and Analytics’ installation identifier is deleted from your device, so that what is collected afterwards is not joined to what came before. Your answer is stored only on this device: on a new one or one restored from a backup you are asked again. Legal basis: your consent (Art. 6.1.a GDPR and Art. 22.2 LSSI).
• Crash logs (Firebase Crashlytics). When the app crashes or an operation gives an unexpected error: device model, Android and app versions, the trace and the technical reason for the error and an installation identifier. They are not associated with your account. Legal basis: legitimate interest in fixing errors. You can turn them off in Settings → “Crash reports” (right to object, Art. 21 GDPR).
• ML Kit diagnostics. The cropping and straightening of labels and the reading of QR codes are done on your device with Google’s ML Kit, which does not send those images to Google. The already cropped image is the one sent to the artificial intelligence (section 3) and the one suggested to you as the coffee’s photo. The captures that the scanner leaves in the app’s temporary memory are deleted when they are more than seven days old, the next time you open it. ML Kit sends Google diagnostic data (device model and version, app version, an identifier per installation and performance and error metrics). Legal basis: legitimate interest in the proper functioning of those features.
• Device identifiers (Firebase and App Check). Firebase assigns an installation identifier, and App Check verifies with Google Play that requests come from the authentic app on an unmodified device, to protect the service from abuse. In the private panel on our website from which roaster applications are reviewed (coffeezapp.com/admin), which only the data controller uses, App Check relies on Google’s reCAPTCHA Enterprise, which analyzes the browser of whoever opens it. Legal basis: legitimate interest in the security of the service.
• Server technical logs (Google Cloud). Google Cloud records every request to our cloud functions, with the time, the IP address, the type of device or browser and the result, also when the website counts the opening of a QR. Some logs also carry your user identifier: the plan change after a purchase, the actions on your roaster application that go through our functions, yours from the app (such as submitting the form or requesting or confirming the code) or the controller’s (in theirs, with the administrator account that did it), the deletion of your account (with what RevenueCat answered when asked to delete your purchase record), the creation of your seal when you publish your first package (only if we approved you before the panel existed) and certain errors, such as a failure when deleting the account. We use them only to maintain the service, fix errors and detect abuse. Legal basis: legitimate interest in the security and proper functioning of the service.

In the development versions of the app, Analytics and Crashlytics are turned off.

3. Artificial intelligence

Label scanning, the recipe generator and the cupping adjustment use Google’s paid Gemini API. We send it only what each feature needs: the label’s image or PDF; the coffee’s data (origin, variety, process, roast, days since roasting and its notes or flavor profile), the recipe and your grinder; and what you describe about the result in the cup. We do not send it your name, your email or your user identifier.

Our servers pass that data to Gemini and do not store it: for each call we only note one use in your counter (section 4) and its cost in a monthly total that is not linked to your account. In earlier versions of the app, the artificial intelligence also estimated the roast from a photo of the beans, which was treated the same way; the current version no longer does so.

Google acts as a processor. With the paid API it does not use what we send it or its responses to train or improve its products; it keeps them for a limited time solely to detect abuse of its usage policies, and may process them in any country where it has facilities (see section 6).

We use the Gemini API on its paid tier, with prepaid balance and a key that only our servers have: the app never connects to Gemini directly. If Gemini is not available, the feature simply does not respond and nothing is stored. Legal basis: performance of the contract (Art. 6.1.b GDPR), because it is the service you ask for when using those features.

4. Control of artificial intelligence use

To share the cost of the service fairly we keep a usage counter per account. If you delete your account we keep for 35 days an irreversible cryptographic fingerprint (HMAC-SHA256) of your email address, which does not allow recovering the email or identifying you, for the sole purpose of preventing the free limit from being evaded by deleting and recreating the account. After 35 days it is deleted automatically, normally within the following 24 hours. Legal basis: legitimate interest in preventing service fraud (Art. 6.1.f GDPR).

5. Providers

These providers process data on our behalf, as processors:

• Google: Firebase (authentication, database, photo storage, cloud functions, Analytics, Crashlytics and App Check), Google Cloud (including the server technical logs and reCAPTCHA Enterprise), the paid Gemini API and ML Kit.
• RevenueCat, Inc. (United States): manages the purchases made on Google Play, associated with your user identifier, and tells us what you have active.
• Mailjet (Sinch Mailjet SAS, France; servers in the European Union): sends the emails with the roaster verification code. It receives the destination address, the subject and the text of the email (the code, your website or your Instagram account and your account’s email masked), without open or click tracking. Mailjet stores the destination address as a contact, with the log of the sends, while the application is pending: when we decide it or you delete the account, we ask it to delete it (if the request fails, we delete it by hand), and Mailjet anonymizes it immediately and deletes it completely within 30 days. Its sending statistics are kept without the address.

The emails you write to us at the coffeezapp.com addresses are received by Cloudflare’s email forwarding service (Cloudflare, Inc., United States), which forwards them to us; those to soporte@coffeezapp.com, also your replies about the roaster application, arrive in a CoffeEZ Gmail (Google) mailbox.

Payments are processed by Google Play under its own privacy policy.

6. Where data is stored and international transfers

The database (Firestore, in the European multi-region eur3, in Belgium and the Netherlands), the cloud functions and the photos of your coffees are on Google servers in the European Union; the functions and the photos, in Belgium. The roaster verification code emails are sent by Mailjet from servers in the European Union.

Some services may process data outside the European Economic Area: authentication, Analytics, Crashlytics, App Check (and reCAPTCHA Enterprise, in the website’s administration panel), ML Kit diagnostics and the Gemini API, which are global Google services; the server technical logs, which Google Cloud stores in its global location, with no guarantee that they remain in the European Union; RevenueCat, which stores data in the United States; Cloudflare’s email forwarding, a United States company, for the emails you write to us; and the Gmail mailbox in which we receive those to soporte@coffeezapp.com, which is a global Google service. Transfers to Google LLC are covered by its adherence to the EU-US Data Privacy Framework and by the European Commission’s standard contractual clauses; transfers to RevenueCat, by the standard contractual clauses; and transfers to Cloudflare, by its adherence to the EU-US Data Privacy Framework and by the standard contractual clauses.

7. Advertising

CoffeEZ does not show ads and does not use Android’s advertising identifier: the app does not request that permission, and Analytics is configured not to collect it nor to use ad personalization signals.

8. Sharing recipes

If you share a recipe through a link or QR code, we store a copy of its content: name, method, quantities, grind, temperature, steps, the notes you wrote in the recipe, the name of the coffee it is associated with and, if you adjusted it by cupping, the latest diagnosis and the proposed change. To share a recipe you need an account with a verified email. To open it the app is enough, even if you have no account. Only whoever has the link or the code can see it: it cannot be searched or listed, and it does not include your name, your email or your user identifier. Before sharing, check that the notes do not contain anything you do not want to show. Links expire after 7 days, or after 180 days if you are a verified roaster; from that moment they stop working and are deleted automatically, normally within the following 24 hours.

9. Retention

Your coffees, recipes and photos (the artificial intelligence proposals, only if you gave the consent in section 13), your roaster application (with the internal notes and the history, also if it was rejected or we withdrew your verification, with its conversation until you apply again, and with the verification code data: the full email address only while it is pending, masked afterwards), your profile (with your plan, your interest in the plans and the roaster you arrived through), your artificial intelligence usage counter and the record of your consent (section 13) are kept for as long as you keep the account. The anonymous session in section 1, until 90 days pass without using the app. Crash reports are kept for 90 days. In Analytics, event data is kept for 2 months, and data linked to the installation identifier, for 14 months from your last activity: the period starts again every time you use the app. When it expires, Analytics deletes it in its monthly cleanup; only aggregate reports remain, which identify no one. Server technical logs are kept for 30 days. The abuse-control fingerprint, 35 days. Shared links, as indicated in section 8. A roaster’s seal, for as long as they keep the account and the verification. The packages they published, with their opening count, for as long as they keep the account, unless we delete them when withdrawing the verification; their batches, the activity record of their catalog and their statistics, for as long as they keep the account. RevenueCat keeps the record of your purchases, associated with your user identifier, for as long as you keep the account; when you delete it, we ask it to delete it (section 11). Mailjet keeps the address to which we send each verification code while your application is pending; when it is decided or when the account is deleted, we ask it to delete it (section 5). The emails in the support Gmail mailbox (section 5), that is, those you write to us at soporte@coffeezapp.com (also your replies about the roaster application) and those we send you from it, are kept for up to 12 months after the application is decided (or after you delete the account, if it was still pending) or after the inquiry is handled; the controller deletes them by hand, with no automatic deletion. The random mark in the per-domain code record expires at the latest 2 days after the day you requested the code and is deleted automatically, normally within the following 24 hours (section 2).

The database has backups: it can be recovered as it was at any moment of the last 7 days, and a daily copy is made that is kept for 7 days. Photos, when deleted, may first go through the storage’s trash. For that reason, what is deleted, also when deleting the account, may remain for up to 7 days in those backups or in the storage’s trash, and then disappears on its own.

10. Your rights

You can access, rectify, erase, restrict and object to the processing, as well as request the portability of your data, by writing to privacidad@coffeezapp.com. You can also lodge a complaint with the Spanish Data Protection Agency (www.aepd.es).

11. How to delete your account and your data

From the app itself: Settings → Delete account. The following are deleted immediately and irreversibly: your coffees, your recipes, your photographs, your roaster application (with the copy of the conversation it keeps, the internal notes and the history), your profile (with your plan, your consent, your interest in the plans and the roaster you arrived through), your usage counter and your access account, both from the device and from our servers. If you are a roaster, also your seal, your batches, the activity record of your catalog, your statistics, the packages you published with their QR code (with the name of your seal) and their opening count; QR codes already printed will stop working. We also remove the reference to you from the profile of those who arrived through your QR codes (section 2). The copies of your packages that other users saved in their pantry, with the name your seal had when they saved them, are those users’ data and are not deleted.

When deleting the account we also ask RevenueCat to delete the record of your purchases. That deletion is not immediate: RevenueCat queues it and completes it shortly afterwards. If your roaster application was pending, we also ask Mailjet to delete the addresses to which we sent your verification codes (section 5). Google Play keeps its own records of your purchases and payments, as an independent controller and under its own privacy policy.

Deleting the account does not cancel a Catador subscription, just as uninstalling the app does not: the subscription is in your Google Play account and would keep renewing. If you have it active, cancel it first in the Google Play app → Payments & subscriptions → Subscriptions → CoffeEZ → Cancel subscription, or at play.google.com/store/account/subscriptions. From CoffeEZ, the “Manage subscription” button on the plans screen takes you to that page.

Deleting the account does not delete data that is not part of it: analytics and crash data, tied to an installation identifier; the recipe links you shared, which do not identify you and expire on their own (section 8); the server technical logs (some with your user identifier), which are deleted when their period ends (section 9); if you requested a roaster verification code, the random mark in the per-domain code record, which can no longer be related to you and expires on its own (section 2); and the emails in the support Gmail mailbox (those you wrote to us at soporte@coffeezapp.com and those we sent you from it), which we delete by hand when their period ends (section 9). After deleting the account, the data in the database and the photos may remain for up to 7 days in backups or in the storage’s trash, and then disappear on their own (section 9).

You can also delete individual data without deleting the account: each coffee from the pantry (together with its photo) and each recipe from its card. To delete other specific data without deleting the account, write to us at privacidad@coffeezapp.com saying which, and we will delete them within a maximum of 30 days.

If you no longer have the app installed, write to us at privacidad@coffeezapp.com stating the email address you signed up with. These requests are not automatic: the data controller handles them by hand. If you write to us from another address, we will ask you to confirm it from your account’s. Then we delete, from the Firebase, Google Cloud, RevenueCat and Mailjet consoles, the same that the app deletes (see above), and we reply when it is done, within a maximum of 30 days. As explained in section 4, after the deletion we keep for 35 days an irreversible fingerprint of your email that does not allow identifying you.

If you are a roaster and want to withdraw a published package without deleting the account, ask at privacidad@coffeezapp.com and we will withdraw it within a maximum of 30 days; its QR code will stop working. Archiving a batch in the app does not withdraw its package: the printed QR keeps working.

12. Minors

CoffeEZ is aimed at people aged 18 or over. We do not knowingly collect data from minors; if you believe a minor has given us data, write to us and we will delete it.

13. Improving the recommendations (optional)

In Settings → “Help improve CoffeEZ” you can give us your consent to use your recipes to improve CoffeEZ’s recommendations for all users. We ask you inside the app and only if you have an account; refusing does not limit any feature.

What data it is: the parameters of your recipes (method, dose, water, grind, temperature and times), their rating from 1 to 5 and what the artificial intelligence proposed to you versus what you finally saved. They do not include your notes or what you write when describing the result in the cup: that is only sent to the artificial intelligence to obtain its answer (section 3) and is not stored. The artificial intelligence proposals are stored on your device, but are only uploaded to your account if you give the consent. When they are used, it will be without your name or your email, and unlinked from your account. They are not sent to Google to train its models.

For now we do not use this data: the improvement process does not exist yet. When it exists, it will only take into account that of the accounts with consent in force.

Legal basis: your consent (Art. 6.1.a GDPR). We store which version of this text you accepted and when. You can withdraw it at any time in Settings → “Help improve CoffeEZ”; from that moment we stop using your data and delete from your account the artificial intelligence proposals stored for this (on your device they stay, in case you give it again). What has already been incorporated anonymously into a recommendation model cannot be separated from it. This data is deleted with your account.

14. Changes

If we modify this policy we will update the date in the heading and, if the change is significant, we will notify you inside the app.