1. Introduction

OSHIYON (hereinafter "the App") is a personal fan activity recording app provided by an individual developer (hereinafter "the Developer"). This Privacy Policy explains how your information is handled within the App.

Contact: support@afropunx.com

2. Information We Collect

The App follows a local-first design. Routine recording, viewing, and statistics are handled on your device, while user-selected research, cloud analysis, image analysis, URL fetching, backup, purchases, and speech recognition may involve external services. The routes and conditions are described in this policy.

2.1 Data You Enter

Inputs, photos, URLs, and speech-recognition results are handled locally and may be sent to cloud features only when you explicitly start them. The App does not collect your entire photo library without permission.

• Oshi information (stage names, character names, SNS account names, profile images)
• Activity logs (text, dates, mood, costs)
• Event information (event names, venues, seats, costs)
• Expense records (amounts, categories, currencies)
• Goods, releases, media appearances, and trip records
• Photos and videos (selected from your library or taken with your camera)
• Titles, descriptions, body excerpts, image text, and similar content obtained from shared URLs, SNS posts, or web pages when you confirm and save them as records

The photo-event candidate feature shows candidates for the relevant time window on your device after permission. It does not save or transmit candidate photos without your confirmation.

2.2 Data Obtained Through Google Sign-In

Google Sign-In is used when you choose cloud backup or another Google-linked feature. The following information may be handled:

• Email address
• Display name (if set)

This information is stored locally on your device for session management and is deleted upon sign-out or after a successful Google disconnect. Ordinary sign-out ends the device session but does not guarantee revocation of the OAuth grant; a successful disconnect is a separate operation. Purchase state, receipts, and the App User ID are handled separately through App Store/Google Play/RevenueCat purchase linkage.

2.3 Device Information

• During cloud backup: Device model name (for backup identification purposes)

2.4 Information Collected by Third-Party SDKs

• Google ML Kit: May collect device information and performance diagnostic data for on-device text recognition (OCR) processing (Google ML Kit Terms of Service)

2.5 Information the App Does Not Ask You to Enter

The App does not ask you to enter the following information in its own screens. Vendor SDKs and external services may have separate processing conditions.

• Legal names, street addresses, phone numbers, driver's license, passport, or other identity documents
• Credit card numbers or bank account information (the App does not receive these payment credentials; App Store / Google Play process the payment, while purchase state and receipt/entitlement may be handled separately)
• GPS coordinates or precise location as an App feature. Third-party SDK processing follows the relevant provider conditions
• Medical information, religious beliefs, political opinions, racial/ethnic origin, sexual orientation
• Contacts (address book)
• Crash reports (no Firebase Crashlytics, Sentry, or similar services are used)
• Advertising profiles and cross-service usage tracking (Firebase Analytics, Mixpanel, Amplitude, or similar) are not used. ML Kit diagnostics and purchase/research operation records may still be processed as described in this policy.
• Advertising identifiers (IDFA / AAID are not used; no ad SDKs are integrated)

3. Purpose of Use

4. Legal Basis for Processing (GDPR Article 6)

For users in the EU and EEA, we process data based on the following legal bases:

• Performance of a contract: Providing the App's core functionality (storing and displaying data)
• Consent: Google Sign-In, cloud backup, voice input, camera, microphone and photo library usage, Gemini API, URL sharing, SNS/OGP fetching, and Research AI (internal testing only; §6.5). Inputs for these features are handled when you explicitly start them; provider-side SDK diagnostics and similar processing follow the provider's conditions.
• Legitimate interest: Ensuring app stability and bug fixes

5. Data Storage

5.1 Local Storage

Your device SQLite database is the main storage for records. When you choose Google Drive, Research relay, cloud parse/image analysis, or purchase linkage, the data needed for that route may be transmitted or retained by the relevant external service (see §6.5).

5.2 Cloud Backup (Optional)

Only when you explicitly choose to do so, the on-device database and app image files, including confirmed records derived from shared URLs and ticket images, can be backed up to Google Drive.

• Storage location: Your own Google Drive's private app folder (appDataFolder)
• Encryption: Client-side AES-256 encryption (optional, user-set password)
• Retention: Up to 3 automatic backups and up to 3 manual backups (older backups are automatically deleted)
• API scope: drive.appdata (access limited to the app-specific folder; we do not access your other Drive files)

6. Third-Party Services

The App is local-first. User-provided text, images, and search keys are sent to an external service when you explicitly start the relevant feature. SDK diagnostics, performance data, or purchase SDK initialization and identifier processing may occur under the relevant provider conditions. Research AI routes through a Developer-operated relay, while parse and image analysis are separate explicit cloud routes.

6.1 Google Sign-In / Google Drive API

Used when you choose cloud backup or another supported Google-linked feature. Google's Terms of Service and Privacy Policy apply. Purchase linkage is a separate App Store/Google Play/RevenueCat route, not Google Sign-In.

Backups are stored in your own Google Drive appDataFolder. A backup may include activity records, oshi information, confirmed records derived from shared URLs, image files, and some settings; it is an external copy managed under Google Drive conditions.

• Google Terms of Service
• Google Privacy Policy

6.1.1 In-App Purchases and RevenueCat

When you use or restore an in-app purchase, App Store / Google Play purchase state, receipts/entitlements, and the RevenueCat App User ID are used to provide and restore the purchase feature and manage usage. The App does not receive credit-card numbers or bank-account information. Deletion and retention of purchase state follow App Store, Google Play, and RevenueCat conditions.

6.2 Google ML Kit (Text Recognition)

The OCR operation uses Google ML Kit and the OCR processing itself occurs on your device. A separate, explicit image-analysis route may send selected images, OCR text, and supporting information such as the current date, language, and registered oshi names to an external service. The ML Kit SDK may collect device and app diagnostic or performance data.

6.3 System Speech Recognition

The voice input feature uses the speech recognition service built into your device.

• Android: Google's speech recognition service is used, and audio data may be sent to Google's servers for processing. On some newer devices, on-device processing may be available. See Google's Privacy Policy for details.
• iOS: Apple's speech recognition framework is used. Whether audio is processed on-device or sent to Apple's servers depends on the OS and recognition service conditions. See Apple's Privacy Policy for details.

The App does not save raw audio files or upload them to an arbitrary URL; it handles the text returned by the operating system recognition API. Audio processing by the OS or recognition provider follows that provider's conditions. Recognized text becomes a record or user content only when you review and save it.

6.4 Other Automatic API Calls

The following feature-specific calls contact external services when the corresponding user action occurs. The exchange-rate request below contains no user identifier; the BYOK and Developer-relay routes have the separate content and identifier scopes described below. SDK diagnostics and purchase processing may also occur under provider conditions.

• frankfurter.app (Exchange Rate API): When you enter a foreign currency in expense records, only the currency pair (e.g., USD→JPY) and date are sent to api.frankfurter.app to fetch the latest rate. Only currency codes and dates are transmitted; no user-identifying information is included.
• X (Twitter) / TikTok / Instagram / general web pages (SNS APIs, OGP, and HTML fetch): When you enter a URL or share one from another app, the App may use that URL or post ID to fetch titles, descriptions, body excerpts, thumbnails, image text, and limited linked content. Fetched content is used for AI-input previews and extraction. Raw HTML or entire fetched pages are not stored as-is. Only extracted fields such as event names, dates, or notes are saved locally after you review and confirm them.
• Google Gemini API (your own key / BYOK): Only if you register an API key you obtained yourself in the settings screen, the API key is included in the request header and text/image analysis requests are sent to Google's Gemini API. Content that may be sent includes text you enter, titles/descriptions/body excerpts fetched from shared URLs, selected or shared images, ticket images, and supporting information needed for image analysis such as the current date, language, and registered oshi names. The API key is obtained and managed directly between you and Google; the App developer is not involved. You can revoke this at any time by removing the key in the settings screen. For Gemini's terms and privacy policy, see Google AI Gemini API Terms of Service. This describes the path that uses your own key (BYOK). When using Research AI with a key provided by the Developer (§6.5; internal testing only), search keys are likewise sent to Google's Gemini API using the Developer's key.

6.5 Research AI (via a Developer-Operated Server)

This feature is currently limited to internal testing by the Developer and is not included in the publicly released app. Before general release, this policy will be updated prior to the feature becoming available.

The Research AI feature (looking up information about your oshi or events, and drafting setlists) requires the internet by nature. Separately from the path where you use your own API key (BYOK; §6.4), to let you use it with a key provided by the Developer, requests route through a relay server operated by the Developer (on Cloudflare Workers). This server also manages usage for this path (balance, claiming, and consumption).

What is sent to this server

• Search keys (text such as the name of the oshi or event you want to look up). These search keys are forwarded downstream to Google's Gemini API using the Developer's key; their handling follows Gemini's terms and privacy policy.
• An ID to distinguish users (the RevenueCat App User ID) and request identifiers.

For the basic oshi/event/setlist relay routes, the server accepts search keys, the RevenueCat App User ID, and request identifiers, and rejects photo attachments. This scope does not apply to the separate parse and analysis/image routes: after consent, parse may receive post text, URL extracts, or OCR text, and analysis/image may receive selected images, OCR, and supporting information such as the current date, language, and registered oshi names.

What this server stores (only as needed for feature and billing management)

What this basic relay does not retain as an independent raw log: search-key text, full AI prompt text, retrieved source text, or general app interaction logs. Search results, per-user operation/consumption records, balance and tombstone data are retained as shown above. Inputs and results for parse/image follow those separate routes.

Stored data is encrypted at rest via Cloudflare's standard feature (AES-256), but the encryption keys are managed by Cloudflare. Data is processed on Cloudflare's global infrastructure (§10).

Deletion requests: Deletion of this server-side data is currently handled manually. Please contact support@afropunx.com. However, only the minimal welcome-bonus claim record (above), which prevents claiming the bonus twice, remains even after deletion.

7. Android Automatic Backup

If backup is enabled on your Android device, app data (activity records, settings, and some images) may be automatically backed up to your Google account through Android's built-in backup service. This backup is encrypted and subject to Google's Privacy Policy.

This is separate from the in-app cloud backup feature. You can disable it in Android Settings > System > Backup.

8. Device Permissions

All permissions are requested at runtime when first needed. You can revoke them at any time in your device settings.

9. Data Security

• Data is stored in your device's local storage
• All cloud communications are encrypted via HTTPS (TLS)
• Cloud backups can be encrypted with AES-256 (user-set password)
• Google Sign-In authentication tokens are not permanently stored on the device
• Data stored by the Research AI relay server (§6.5; internal testing only) is encrypted at rest via Cloudflare's standard feature (AES-256), but the encryption keys are managed by Cloudflare

10. International Data Transfers

Data may be transferred outside your country in the following cases:

• Google Drive backup (Google's global infrastructure)
• Android automatic backup (Google's global infrastructure)
• Speech recognition on Android (Google's servers)
• Speech recognition on iOS in some circumstances (Apple's servers)
• Gemini API if you configure an API key (Google's global infrastructure)
• Shared URL, SNS, and OGP fetching (the services hosting URLs you enter or share)
• The Research AI relay server (internal testing only; Cloudflare's global infrastructure) and its downstream Google Gemini API (via the Developer's key; Google's global infrastructure)

Japan has received an adequacy decision from the European Commission (January 23, 2019, reaffirmed in 2023), allowing personal data transfers from the EU/EEA to Japan without additional safeguards. Transfers to Google, Apple, and other external services are subject to each provider's privacy policy and transfer safeguards such as Standard Contractual Clauses (SCCs).

11. Data Retention and Deletion

OSHIYON is local-first, and on-device records are subject to the App's deletion procedures. Copies in Drive, App Store/Google Play and RevenueCat data, relay records, and parse/image service data follow each service's procedure, manual request, or designed TTL. Values such as 24 hours and 60 days are design values and do not guarantee a measured completion time for every deletion. Research AI server-side data is handled as described in §6.5.

11.1 Deletion Procedures (Individually Available In-App)

Each of the following can be run separately in the App. External copies and server records remain subject to the relevant service deletion procedure and retention conditions.

• Delete individual records: Each record screen → Delete button
• Delete all local data: Settings → Data Management → Delete all local data
• Delete cloud backup: Settings → Cloud Backup → Delete backup
• Sign out from Google: Settings → Cloud Backup → Sign Out (deletes the on-device account cache; it does not guarantee revocation of the OAuth grant)
• Disconnect Google account: Settings → Data Management → Disconnect Google Account (only after OAuth revocation succeeds, the connection state and on-device account cache are cleared; on failure, the connected state is retained so you can retry)
• Delete all app data: Uninstall the app (all on-device data is deleted by the OS)

11.2 Data Types Deleted and Retention Periods

On-device records are subject to the App's deletion procedure. Drive copies, App Store/Google Play and RevenueCat purchase data, Research relay records, and parse/image service data follow each service's processing conditions, manual requests, or designed TTL. The basic relay does not independently retain raw search-key, prompt, or source text as a log, while results and operation records described in §6.5 may still be retained. Values such as 24 hours and 60 days are design values and do not guarantee a measured completion time for every deletion.

11.3 If You Have Already Uninstalled the App

Uninstalling the app deletes all on-device data, but if you previously used cloud backup, the backup files remain in your Google Drive. To delete them:

1. Visit drive.google.com/drive/settings
2. Open the "Manage apps" section
3. Select "OSHIYON"
4. Click "Delete hidden app data"

11.4 Additional Support for Deletion Requests

If you encounter issues with the procedures above or need additional support, contact us at:

• Email: support@afropunx.com
• GitHub Issues: github.com/zrockafropunx/oshiroku/issues

On-device records are deleted through the App. For Drive copies, purchase linkage, relay data, and parse/image service data, use the relevant service procedure or contact support@afropunx.com. Research relay data is handled manually as described in §6.5.

12. Your Rights

Under applicable laws (Japan's Act on the Protection of Personal Information, GDPR, etc.), you have the following rights:

• Right of access: Access your data (directly viewable within the App)
• Right to rectification: Correct your data (directly editable within the App)
• Right to erasure: Delete your data (using the App's delete function or by uninstalling; server-side data of the internal-testing-only Research AI feature is deleted upon request to the email contact, see §6.5)
• Right to data portability: Export your data (Settings > Data Management > JSON/CSV Export)
• Right to restrict processing: Stop specific features (revoke permissions, disable backup, etc.)
• Right to object: Object to data processing (contact support@afropunx.com)

The App does not perform automated decision-making or advertising profiling. OCR and speech recognition help with data entry; explicit cloud analysis may process the selected input to produce results.

13. Withdrawing Consent

You can withdraw your consent at any time through the following methods:

• Google Sign-In authorization: Settings > Data Management > Disconnect Google Account (if it fails, the App keeps the connected state so you can retry while online, or remove access from your Google Account's connected apps)
• Camera, Microphone, Photo Library: Device Settings > Apps > OSHIYON > Permissions
• Notifications: Device Settings > Notifications > OSHIYON
• Android Automatic Backup: Device Settings > System > Backup

Withdrawing consent does not affect the lawfulness of processing carried out before the withdrawal.

14. Data Sharing with Third Parties

The App does not use data for advertising or cross-service tracking. Some SDKs may provide diagnostic or performance data, and purchase/research operations use identifiers needed for those features. We do not sell or share your data for advertising or marketing purposes.

However, data may be processed through external services in the following cases. User-selected features are listed below; SDK diagnostics may also follow the provider's conditions.

• When using cloud backup (Google Drive)
• When using voice input (system speech recognition service)
• When using the Gemini API (only if you configure an API key)
• When using shared URL, SNS, or OGP fetching (only when you enter or share a URL)
• When using Research AI, parse, or image analysis (a relay, Google Gemini API, or other stated provider; see §6.5)

15. Children's Privacy

The App is not intended for children under 13. We do not knowingly collect data from children under 13.

16. Right to Lodge a Complaint

Users residing in the EU/EEA have the right to lodge a complaint with their local data protection supervisory authority regarding the processing of personal data (GDPR Article 77).

17. Changes to This Policy

This Privacy Policy may be updated as necessary. If significant changes are made, we will notify you through updates to this page and app updates.

18. Contact Us

For questions or requests regarding this Privacy Policy, please contact us at:

Email: support@afropunx.com