Privacy policy

Last updated: October 9, 2026

This policy explains what the Sezgi app collects, why, who it is shared with, how long it is kept, and what you can do about it. It covers the Sezgi app for Android and iOS and the services behind it.

1. Who runs Sezgi and how to reach us

Sezgi is operated by OZCANLAR (Ozgur Ozcan, sole proprietor), Toronto, Ontario, Canada. We are the controller of the personal data described in this policy.

Email for privacy requests, questions and complaints: privacy@yinyangstudio.app

The person responsible for protecting personal information (including under Quebec's Law 25 and Canada's PIPEDA) is Ozgur Ozcan, who can be reached at the same address.

2. What we collect

Account

When you first open Sezgi, an anonymous account is created with a random account ID. You do not need an email address or a name.
If you choose to sign in with Google or with Apple, we receive the email address and, if the provider shares it, the name of that account. Apple may give us a private relay address instead of your real email.

What you give us for readings

Text you type: questions, dream descriptions, names and similar details.
Birth details, if you add them: date, time, city and the coordinates and time zone we derive from that city. They are used to work out your sign and chart, and some readings use your first name and birth date.
Choices you make in the app, such as the cards or runes drawn, the reading type, your focus and preferred tone.
Photos, only when you use a photo reading: a coffee cup, your palm or your face. See section 4 for how photos are handled.

What Sezgi creates for you

Your readings and short summaries of them, which form your journal.
Reading memory: Sezgi uses short summaries of your earlier readings to make new readings continue the story. It is on unless you turn it off in You → Settings → Reading memory.
Credits, streaks, daily usage counts, referral codes and the record of rewarded ads you watched.
If you tell us whether a reading spoke to you: your answer (it landed or it missed), the reason you picked if it missed, the app language, and which AI model wrote that reading.
If you report a reading as harmful or upsetting: a copy of that reading, the reason, the app language and the time, kept so a person can review it.
Your answer that you are 18 or older and that you accept this policy and the terms, the version you accepted, and when you gave it.

Purchases

If you subscribe or buy a pack, Google Play or the App Store handles the payment. We never see your card. We receive the purchase status, product, dates, country and price from the store through RevenueCat.

Device and usage information

Crash reports: device model, operating system, app version and the error, linked to your account ID so we can fix problems you hit.
Usage events (for example which features you open). In the EEA, the UK and Switzerland, where the law requires consent, Google's consent form asks first: we collect usage events only if your choices in that form allow it, and if you decline, analytics stays off. Elsewhere they are collected by default.
Advertising: the ad SDK uses your device's advertising ID and IP address (which gives an approximate location) to deliver and measure ads. For our accounts we also record the estimated revenue of each ad shown to you (format, amount, currency), linked to your account.
A device code for fraud prevention: a one-way code made from your phone's Android ID with a secret key, so it cannot be turned back into the ID. It lets us give welcome and daily credits once per phone rather than once per new account.
A push notification token, if you allow notifications, and your notification settings.
The language of the app and, for daily limits, the date of your activity.
How you found Sezgi, read once on Android when the app first opens: the campaign source and name and the ad click ID that Google Play passes to the app at install, and the invite code of a shared link. It is stored with your profile.

We do not collect your contacts, your precise location, or your microphone. Text-to-speech, where offered, runs on your device.

3. Why we use it, and on what legal basis

Where the EU or UK GDPR (or a similar law) applies, we rely on these legal bases:

Purpose	Legal basis
Creating your account, writing your readings, keeping your journal, credits and purchases	Performance of our contract with you
Reading memory across readings	Contract; you can switch it off at any time
Face readings (photo of your face)	Your explicit consent, given in the app before the first face reading (GDPR Art. 9(2)(a) where the photo counts as special category data)
The device code, and the ad revenue and reading-feedback records	Our legitimate interest in preventing credit abuse, keeping our accounts, and improving the readings
Usage analytics	In the EEA, the UK and Switzerland, your consent in Google's consent form; analytics runs only if your choices there allow it. Elsewhere, our legitimate interest in understanding how the app is used
Personalised ads	In the EEA, the UK and Switzerland, your consent in Google's consent form; ads are personalised only if your choices there allow it, and otherwise they are not. Elsewhere, our legitimate interest in funding the free tier, and you can opt out at any time (see section 9). Ads on iPhone are always non-personalised.
Push notifications	Your consent through the system permission
Crash reports, security, preventing abuse of free credits and rewarded ads, keeping a record of account deletions	Our legitimate interest in a working, safe and fair app
Tax and accounting records of purchases	Legal obligation

We do not use your data to make decisions that have legal or similarly significant effects on you. Readings are generated by AI for entertainment and reflection only.

4. Photos

Photos are sent from the app to our server and on to the AI provider that writes the reading. They are not saved in our database or file storage; only the reading text and the fact that a photo was used are saved.
Palm and face photos go only to Anthropic. Coffee cup photos go to Moonshot AI or to Anthropic. The journal summary of a photo reading is also written by Anthropic.
Anthropic deletes the photos it receives within 30 days, unless it flags one under its usage policy. Moonshot AI may keep a coffee cup photo while our account with it is active and may use it to improve its models.
With reading memory on, the one-sentence summary of an earlier reading, including a photo reading, can be part of a later reading that another provider writes. The photo itself never is.
The face reading does not identify you, verify who you are, or build a face template. It produces a written interpretation only. We ask for your explicit consent before the first face reading. To withdraw it, stop using face readings or delete your account.

5. Who processes your data for us

We share personal data only with the providers below, so that they can run Sezgi for us. Some of them may also use it for their own purposes, as explained below for Moonshot AI, DeepSeek and Google's advertising. We do not sell your personal data.

AI providers that write readings

When you ask for a reading, our server sends the provider the text you typed, the choices you made, the details the reading needs (for some readings your first name and birth details, and with reading memory on, short notes from earlier readings) and, for photo readings, the photo. We do not send your email address or your account ID to AI providers. Which provider writes a reading depends on the reading type, whether you have Premium, the reader you choose, tests we run, and which providers are available at that moment.

Besides Sezgi's own voice, you can choose one of six AI readers for a text reading. Each reader is one provider's model, and a reading you give to a reader goes only to that provider: the Lamp to Anthropic (Claude), the Swallow to Moonshot AI (Kimi), the Quill to DeepSeek, the Key to Z.ai (GLM), the Hourglass to Google (Gemini; it reads only for people 18 or older) and the Compass to OpenAI (GPT). Readers read text traditions only; photos never go to a reader you choose.

Provider	What it receives	Where
Anthropic, PBC (Claude)	Premium text readings, all palm and face readings, the backup for coffee cup readings, the second choice when DeepSeek cannot write a free text reading, a new version of a Premium or photo reading that failed our quality check, and readings you give to the Lamp. Also writes the journal summaries of photo, combo, deep and Premium readings, and of other readings when DeepSeek is unavailable.	United States, where Anthropic stores data; a request may also be processed in Europe, Asia or Australia
Hangzhou DeepSeek Artificial Intelligence Co., Ltd. (DeepSeek)	Free text readings other than tarot and dreams; readings you give to the Quill; the text of finished free text readings (such as tarot, dream or crystal ball readings) to write the short summary for your journal; and the daily horoscope, which contains no data about you. Never photos. DeepSeek does not write the summaries of photo, combo, deep or Premium readings; Anthropic does.	People's Republic of China
Moonshot AI Pte. Ltd. (Kimi)	Free tarot and dream readings; a new version of a free text reading that failed our quality check; free text readings when both DeepSeek and Anthropic fail; readings you give to the Swallow; coffee cup photos with the related prompt; journal summaries of free readings when DeepSeek and Anthropic are unavailable; and the daily horoscope when DeepSeek fails (no data about you).	Singapore
Jingsheng Hengxing Technology Pte. Ltd. (Z.ai, GLM models)	Readings you give to the Key and, in limited tests, some free tarot and dream readings and some Premium text readings. Text only, never photos.	Singapore, where Z.ai says it generally processes API data; companies in its group may also handle it
Google Cloud Canada Corporation (Gemini API)	Readings you give to the Hourglass. Text only, never photos.	Any country where Google or its agents have facilities
OpenAI, LLC (GPT models)	Readings you give to the Compass. Text only, never photos.	United States, where OpenAI processes API requests by default

Anthropic, Z.ai, Google and OpenAI process what we send them under data processing terms and say they do not use it to train their models. Google keeps what it receives for up to 55 days and OpenAI for up to 30 days to detect abuse. Moonshot AI's terms allow it to use the content it receives to improve its services, including its models, unless agreed otherwise in writing, and DeepSeek's terms for developers make no commitment either way. That is one reason we never send AI providers your email address or account ID, and why palm and face photos go only to Anthropic.

Other service providers

Provider	Purpose	Where
Supabase	Database, sign-in and server functions	Database in Frankfurt, Germany (AWS eu-central-1); server functions run in the region nearest to you
Google: Firebase Crashlytics	Crash reports	United States and other Google locations
Google: Firebase Analytics	Usage analytics (with your consent where required)	United States and other Google locations
Google: Firebase Cloud Messaging	Delivering push notifications	United States and other Google locations
Google: AdMob and the User Messaging Platform	Showing ads to free users and asking for consent. Google acts as an independent controller for some advertising processing; see how Google uses data from partner apps.	United States and other Google locations
Google Sign-In, Sign in with Apple	Optional sign-in	United States
RevenueCat, Inc.	Managing subscriptions and packs. RevenueCat receives your account ID and purchase records from the store.	United States
Google Play, Apple App Store	Payments, refunds and app delivery, under their own privacy policies	Worldwide

We may also disclose data if the law requires it, to protect people's safety, or to a successor if Sezgi is sold or transferred, in which case this policy continues to apply.

6. International transfers

Sezgi is run from Canada and our database is in Germany. The AI providers in section 5 are in the United States, Canada, the People's Republic of China and Singapore; Anthropic may also process a request in Europe, Asia or Australia, and Google in any country where it or its agents have facilities. Our other service providers are mainly in the United States. These countries may not offer the same protection as yours.

Where the law of your country requires a transfer mechanism, Anthropic and OpenAI provide data processing terms that include the European Commission's standard contractual clauses, with the UK addendum; Google and Z.ai provide data processing terms for API customers. DeepSeek and Moonshot AI do not publish data processing terms. For every provider we limit what it receives as described in section 5, we never send an AI provider your email address or account ID, and all transfers are encrypted in transit.

7. Storage and security

Our database encrypts data at rest and all connections use TLS. Access rules in the database let each account read only its own data.
On your device, Sezgi keeps settings, a cache and readings you saved in the app's private storage. The app does not add its own encryption to this storage; on a phone that is not rooted or jailbroken, other apps cannot read it. It is removed when you delete your account, clear the app's data or uninstall the app.
No method of storage or transfer is completely secure.

8. How long we keep data

Account, profile, credits and settings: until you delete your account.
Readings and their summaries: until you delete them with your account, and at most 24 months after each reading, when they are deleted automatically.
Photos: not stored by us; our server holds a photo only while the reading is written. The AI provider that receives it keeps it under its own terms (see section 4).
Usage and cost logs for readings, rewarded-ad records, reading feedback and push tokens: until you delete your account (feedback on a reading also goes when that reading is deleted).
Readings you report as harmful or upsetting: kept for review for 90 days after you report them, then deleted (sooner if you delete your account), even if you delete the reading itself.
The device code: kept after you delete your account, no longer linked to it, so that a phone cannot collect welcome credits again with every new account. Ad revenue records are kept for our accounts with your account ID removed.
Record of an account deletion (the former account ID, time, and the IP address of the request): kept for 24 months after the deletion to show that we handled the request. The IP address is removed from it after 90 days.
Database backups: deleted data can remain in encrypted backups for a short time until they are overwritten.
Crash reports: kept by Firebase Crashlytics for up to 90 days. Analytics events: kept for the retention period set in our Google Analytics account, at most 14 months.
Purchase records held by RevenueCat and the stores: kept by them as long as needed for billing, tax and fraud prevention. Deleting your Sezgi account does not delete these records automatically; ask us and we will request deletion from RevenueCat.

9. Your choices and rights

In the app

Export your data: You → Settings → Export my data.
Delete your account: You → Settings → Delete account. You can also ask us to delete it without the app; see how to delete your account.
Turn reading memory off: You → Settings → Reading memory.
In the EEA, the UK and Switzerland, change or withdraw your choice for analytics and personalised ads at any time in You → Settings → Privacy choices, or Settings → How ads work → Manage ad consent. Your new choice applies from then on. Anywhere, you can limit personalised ads by resetting or deleting your device's advertising ID in your phone's settings. Turn notifications off in Settings or in your phone's settings.

To use any right below, email us at the address in section 1. We answer within 30 days, or sooner where the law requires. We may need to confirm that the account is yours, for example by asking you to send the request from the app. You will not be treated differently for using your rights.

EU, EEA, UK and Switzerland

You have the right to access, correct, delete, restrict and port your data, to object to processing based on legitimate interests, and to withdraw consent at any time (without affecting earlier processing). You can complain to the data protection authority where you live or work.

United States

Depending on your state (including California, Colorado, Connecticut, Virginia, Utah, Texas, Oregon and others with similar laws), you can ask to know, access, correct and delete your personal data, and to get a copy. We do not sell personal data. Personalised advertising through AdMob may count as "sharing" or "targeted advertising"; you can opt out by deleting or resetting your device's advertising ID or turning off ads personalisation in your phone's settings, through a privacy choice in the app where one is shown, or by emailing us. Face photos are used only to write the reading you ask for and are not stored. If we refuse a request, you can appeal by replying to our answer.

Canada (including Quebec)

Under PIPEDA and Quebec's Law 25 you can access and correct your personal information, withdraw consent, and ask for your data in a structured, commonly used format. You can complain to the Office of the Privacy Commissioner of Canada or, in Quebec, the Commission d'accès à l'information. The person responsible for personal information is named in section 1.

Brazil

Under the LGPD you can confirm whether we process your data, access, correct, anonymise, block, delete and port it, learn who we share it with, and withdraw consent. You can complain to the ANPD.

Türkiye

Under KVKK (Law No. 6698), Article 11, you can learn whether your data is processed, ask for information about it, learn the purpose and the recipients in Türkiye and abroad, ask for correction or deletion, object to results produced solely by automated analysis, and claim compensation for unlawful processing. You can complain to the Personal Data Protection Authority (KVKK).

India

Under the Digital Personal Data Protection Act, 2023, you can ask for a summary of your data and who it is shared with, correct, complete or erase it, withdraw consent, nominate someone to act for you, and raise a grievance with us at the address in section 1. If you are not satisfied, you can approach the Data Protection Board of India.

Japan

Under the APPI you can ask us to disclose, correct, add to, delete, or stop using or providing your data. Some providers in section 5 are in foreign countries; you can ask us for information about their data protection systems and the measures they take.

10. Adults only

Sezgi is for adults only. You must be 18 or older to use it, and the app asks you to confirm this before your first reading. It does not accept a birth date under 18. We do not knowingly collect data from anyone under 18. If you think someone under 18 has used Sezgi, contact us and we will delete the account.

11. Changes to this policy

When we change this policy we update the date at the top. If a change is significant, for example a new kind of data or a new provider, we will tell you in the app before it applies.

12. Contact

Questions, requests and complaints: use the contact details in section 1.