Intou Privacy Policy

Last updated: 23 October 2026

This policy explains what data the Intou app (Android and iPhone) processes, and how. I, Eyay, build and publish Intou, and I am responsible for the data described here (the “controller” under the GDPR and the “veri sorumlusu” under Turkey’s KVKK). Questions and requests: merhaba@eyay.space.

The short version

• Your full chat never leaves your phone. The statistics (who writes more, reply times, who starts) are counted on your phone, without a network connection.
• The AI reading is optional and runs only if you agree: up to 32 short, masked excerpts for the preview and up to 96 for the full report. You can read them before they are sent, or skip AI completely.
• My server passes the excerpts to a third-party AI service and never writes them to disk or to logs. A result waits in memory for at most 15 minutes in case you retry.
• No account, no sign-in, no ads, no analytics and no crash reporting tools. Purchases go through Google Play (the App Store on iPhone) and RevenueCat, under a random ID instead of your name.
• Your reports live on your phone. You can delete one or all of them in the app at any time.

Data collected

These are the data that leave your phone and reach me, or the service providers working for me, even briefly. The headings follow the data types in Google Play’s Data safety section. None of it is sold, used for ads or used to track you across other companies’ apps or websites.

• Messages (masked chat excerpts): sent only if you agree to the AI reading. Purpose: app functionality, that is the AI preview and the full report. Optional; without AI you still get the statistics.
• App activity, other user-generated content: the mode you choose (We’re flirting, We’re together or We’re friends) and any zodiac signs you pick for the full report. They travel with the AI requests. Purpose: app functionality.
• App activity, other user-generated content (AI result reports): if you report an AI result in the app, the reason you pick, your optional note, where you saw the result (the preview or the full report), whether it was the sample chat, the mode and the language. The app masks both names and any contact details in your note before sending it. Purpose: safety and app functionality, that is reviewing problem AI results and improving how the AI is instructed. Optional.
• Device or other IDs: the random purchase ID the app creates on first launch (“intou_” followed by 32 random characters) and the IP address my server sees when you connect. Purpose: app functionality (purchases and report credits), fraud prevention and security (usage limits), and tracking purchases in the RevenueCat dashboard.
• Financial info, purchase history: which product you bought, when, at what price and in which country. It comes from Google Play to RevenueCat. Purpose: app functionality and tracking purchases. Your card details never reach me or RevenueCat.
• Not collected: your name, email address, phone number, contacts, location, photos and other media, the chat file or the full chat, advertising ID, usage analytics and crash reports.

What happens on your phone

• You export a one-to-one WhatsApp chat “without media” yourself and either send it to Intou from the share menu or pick the file in Intou. Intou reads it on your phone and saves neither the file nor the full chat.
• On Android, a shared file is read without saving a copy. On iPhone, the copy iOS hands the app is deleted right after the app tries to read it, even if that fails. If you pick the file in Intou, the temporary copy the file picker leaves in the app’s cache is cleared too.
• The statistics are calculated on your phone, with no network request: message counts, reply times, who starts conversations, the longest silence, emoji and topics.
• A saved report contains the two names as they appear in the chat, number summaries, topic headings, the AI preview if you used it and, once you unlock it, the full report. If the AI picked one message as the best, that one message is part of the full report. Zodiac signs you pick are saved with the report. The app keeps up to 30 reports; when you save more, the oldest one is removed.
• If you agree to AI, the masked excerpts are saved with the report so you can unlock it later. They are deleted as soon as the full report is ready, when you delete the report, or after 30 days if you haven’t unlocked it.
• The preview’s hidden answers are stored on your phone, encrypted with a key that only my server has. That is how the full report always matches the preview. The server accepts them back for 30 days.
• The sample chat (“Try a sample chat”) runs entirely on your phone and never contacts my server.

What is sent for the AI reading

Before anything is sent, the app tells you how many excerpts will go and lets you read them. If you say no, you keep the statistics made on your phone. The app asks again for every new chat. What is sent:

• Excerpts: up to 32 for the preview and up to 96 for the full report, picked from the chat on your phone. Each excerpt is at most 600 characters.
• Masking: in every excerpt the two names become [you] and [other person], and phone numbers, email addresses, links and @usernames are hidden. Masking is automatic and can miss things such as a nickname or a home address, so please look through the excerpts before you agree.
• Labels instead of names and dates: who sent each excerpt (you or the other person) and a day number (day 1, day 2…) instead of a date.
• A number summary: the total number of messages, how many days the chat covers and, for each person, their message count, the conversations they started and their typical reply time.
• Settings: the language, the mode you chose and, for the full report, the writing tone that goes with the mode and any zodiac signs you picked.
• The random purchase ID: sent to my server for usage limits and to use a credit for a full report. My server does not pass this ID to the AI service.
• For the full report: the preview’s encrypted answers, so the report agrees with the preview.
• Never sent: the full chat, the chat file, the names, your contacts, photos or other media, your phone number or your location.

My server receives the request, sends the excerpts and the number summary to the AI service and returns the result to your phone. It never writes excerpts or results to disk and never logs them, so I have no copy of your chats or reports. To make a retry safe, the result waits in the server’s memory for at most 15 minutes and is then dropped. So that the same request can’t be charged twice, for 24 hours the server keeps only the request’s random ID and a one-way fingerprint (hash) of the request in memory; neither can be turned back into text.

A third-party AI service processes the excerpts on my behalf and sends back the reading. I call it with response storage switched off. It doesn’t use your excerpts to train its models. Only if its automated checks flag possible abuse may it keep samples for a limited time so that they can be reviewed.

The reading is generated automatically and is meant for fun. If the excerpts show signs of threats, pressure, violence or self-harm, or that someone may be under 18, it stays calm and careful instead: it leaves out the playful parts and may suggest talking to someone you trust or to local support. It doesn’t lead to any decision about you with legal or similarly significant effects.

Purchases

• You pay through Google Play (the App Store on iPhone). I never see your card details, your name, your Google account or your Apple Account.
• Purchases and report credits are managed by RevenueCat on my behalf. At RevenueCat, your customer ID is the random purchase ID instead of a name or an email address.
• RevenueCat receives this ID, the details of your purchases from the store (such as the product, date, price and country) and technical details that its software in the app sends with each request: the app version and platform, your device’s brand and model, its system version, your language and region settings and the country of your store account. On Android no device identifier is sent. On iPhone this also includes the identifier for vendor (IDFV) that Apple gives my apps on your device; it is not the advertising ID.
• When you open a full report, my server asks RevenueCat to use one credit. If the report can’t be made, the server gives the credit back. If giving it back fails, the server writes a keyed, one-way fingerprint of the random ID, together with the request ID (never any content), to its log so I can fix it by hand.
• So that unused credits survive a reinstall or a new phone, the random ID is kept in Android’s backup on Android and in iCloud Keychain on iPhone. Nothing else from the app goes into the Android backup.

Other connections and logs

• Start-up check: when the app opens, and before the AI steps, it asks my server whether AI is available. This request carries no personal content.
• IP address and usage limits: every connection shows your IP address to the server you connect to. My server uses it only to answer and to apply usage limits that prevent abuse; for an IPv6 address it counts the whole /64 block, not the single address. AI previews are counted per IP address and per random ID for each day (UTC), full reports for each hour, failed full reports per random ID for each day (UTC), and AI result reports per IP address for each hour. These counters are kept in memory only, never written to disk, and start again from zero when their day or hour is over. The IP address is not used to find your location.
• Server logs: my server writes one short line per request: the request method, the address, the status code, an error code and the duration. That line contains no IP address, no ID and no content.
• AI result reports: if you report an AI result in the app, my server writes it to its logs as one line: the reason, your optional note (with both names and contact details masked), where you saw the result, whether it was the sample chat, the mode and the language. That line contains no IP address, no purchase ID, no chat, no excerpts and none of the AI’s text. I read reports only to review AI results and improve how the AI is instructed. Please don’t put excerpts of the chat or sensitive details about yourself in the note.
• Hosting: my server runs at a cloud hosting provider, in a data centre in the European Union (Sweden). The provider may keep standard technical logs for security and operations. The server’s logs and these technical logs are kept for at most 30 days.
• Sharing: share cards are made on your phone. You choose where they go; they don’t reach me.
• Your phone’s backups: on iPhone, if you use iCloud Backup or back up to a computer, iOS includes your saved reports in that backup; I have no access to it. On Android, reports are not included in backups.
• Support: if you email me, I use your email address and what you write only to answer you, and delete it 12 months after your request is closed. My email is hosted by Natro. Please don’t send me your chat file.

The other person in your chat

A chat always involves someone else, so Intou sends as little about them as possible: the full chat stays on your phone, names are masked, only a small sample is sent and only with your OK, and my server keeps no content. Before AI is used, you confirm that you are allowed to share the chat. Please think of the other person before you analyze or share anything. If you are the other person and have a question, write to me; because I keep no chats, excerpts or names, I hold no record about you.

Purposes and legal bases

• AI preview and full report: your consent, given in the app before anything is sent (GDPR Art. 6(1)(a); where excerpts may reveal sensitive information such as health or sex life, your explicit consent under Art. 9(2)(a); KVKK Art. 5(1) and 6). For the other person’s words in the excerpts: my legitimate interest in providing the analysis you ask for, with the safeguards above (GDPR Art. 6(1)(f); KVKK Art. 5(2)(f)).
• Purchases, credits, restoring purchases and returning credits for failed reports: performing my contract with you (GDPR Art. 6(1)(b); KVKK Art. 5(2)(c)).
• Usage limits, security and preventing abuse: my legitimate interest in keeping the service available and safe (GDPR Art. 6(1)(f); KVKK Art. 5(2)(f)).
• AI result reports: my legitimate interest in keeping AI results safe and accurate (GDPR Art. 6(1)(f); KVKK Art. 5(2)(f)). Sending a report is optional.
• Answering your emails: my legitimate interest in helping you, or my contract with you (GDPR Art. 6(1)(b) and (f); KVKK Art. 5(2)(c) and (f)).
• Legal obligations: where the law requires it, for example a valid request from an authority (GDPR Art. 6(1)(c); KVKK Art. 5(2)(ç)).

Third parties and service providers

• A third-party AI service: the masked excerpts, the number summary and the settings listed above. It works on my behalf.
• RevenueCat, Inc. (purchase management): the random ID, purchase details and the technical details listed above. It works on my behalf.
• Google (Google Play) and Apple (App Store): they process your payment as the stores, under their own privacy policies.
• A cloud hosting provider: data in transit, my server’s logs (AI result reports included) and technical logs for my server.
• Authorities: only if the law requires it.

I don’t sell data, I don’t give it to advertisers or data brokers, and I don’t use it to train AI models.

Transfers to other countries

My server runs in a data centre in the European Union (Sweden). The AI service may process the excerpts in data centres in different countries. RevenueCat, Google and Apple are US companies, so purchase data may be processed in the United States. Your data may therefore be processed outside your country. Transfers from Turkey follow Article 9 of the KVKK. When data from the EU/EEA or the UK goes to a country without an adequacy decision, I rely on a recognized safeguard such as the European Commission’s Standard Contractual Clauses. Ask me if you’d like details.

How long data is kept

• On your phone: reports until you delete them or the app (up to 30 reports); masked excerpts until the full report is ready, you delete the report, or 30 days pass without unlocking it.
• On my server: excerpts and results in memory for at most 15 minutes; the request ID and its fingerprint in memory for 24 hours; the usage counters in memory until their day (UTC) or hour is over. Excerpts, results and chat content are never written to disk; AI result reports live in the logs below.
• The AI service: nothing kept for training; samples flagged by automated abuse checks may be kept for a limited time for review.
• RevenueCat: while you may still use your purchases and credits and as long as the law requires, or until you ask for deletion where the law allows.
• Logs: the server’s request lines, AI result reports, the rare refund-failure line and the hosting provider’s technical logs for at most 30 days.
• Emails to me: deleted 12 months after your request is closed.

Deleting your data

• One report: open My reports (the bookmark at the top right of the home screen), swipe the report to the left or press and hold it, and confirm.
• All reports: tap the shield (Privacy) at the top right of the home screen, then Delete all reports.
• The app: uninstalling it removes its reports from your phone. On iPhone, backups you made earlier may still contain them until those backups are replaced or deleted.
• Purchase records: the random ID stays in the backup so your credits aren’t lost. To have the RevenueCat records linked to it deleted, email merhaba@eyay.space with the order number from your Google Play (or App Store) receipt.
• My server: there’s nothing from your chat to delete; it keeps nothing beyond the times above. AI result reports are not linked to any ID and are deleted with the logs after at most 30 days.
• Account: Intou has no accounts, so there is no account to delete.

Your rights

Depending on where you live, you can ask me to tell you what data I have about you and give you a copy, correct it, delete it, restrict or object to its use, or send it to you in a portable format. In Turkey you also have the rights in Article 11 of the KVKK, including knowing who your data is transferred to in Turkey or abroad, objecting to a result against you that comes only from automated analysis, and claiming compensation for damage caused by unlawful processing.

You can withdraw your consent at any time; this doesn’t affect what happened before. In practice: choose “Continue without AI” and delete your reports. To make a request, email merhaba@eyay.space; I answer within 30 days, free of charge. For purchase records, include the order number from your store receipt. I may ask for information to confirm that a request is really yours.

If you are not satisfied with my answer, you can complain to the data protection authority where you live or work in the EU/EEA, to the ICO in the UK, or to Turkey’s Personal Data Protection Board (kvkk.gov.tr).

Security

• Connections between the app, my server, the AI service and RevenueCat are encrypted (HTTPS).
• The app holds no AI keys; the AI can only be reached through my server.
• The preview’s hidden answers are sealed with AES-256-GCM encryption.
• Your saved reports are protected by your phone’s own app sandbox and screen lock.
• No system is perfectly secure. If something goes wrong, I will act as the law requires.

Children

Intou is for adults aged 18 and over. It isn’t directed at children, and I don’t knowingly collect data from children. If you believe a child has used Intou, email merhaba@eyay.space.

This page

This policy is published on eyay.space. How the site itself handles data (cookies, visit statistics, server logs) is explained on the eyay.space privacy page.

Changes

When I update this policy, I post it on this page and change the date at the top. If what is sent for the AI reading ever changes, the consent screen in the app will show it before anything leaves your phone.

Contact

Eyay · merhaba@eyay.space. Intou is not affiliated with WhatsApp or Meta.