Kanade

Privacy Policy

Established · Last revised · Publication revision 2026-11-19

This version applies when the corresponding app release becomes available.

Scope and operator

This policy covers data handling in Kanade Play and Kanade Staff, both operated by Masaru Hirose.

Authentication and devices

Both apps use store-provisioned account details, store and role, QR exchange values, a hashed device identifier, and session timestamps and status to sign you in, check permissions, and revoke access. The camera or selected QR image is processed on your device. Only the values needed for sign-in are sent. This uses Cloudflare Workers, tenant-specific Turso, Google Firebase Authentication, and App Check.

Kanade Play

To reduce repetition, store music preferences and each playback account’s track IDs, start times, and sessions are stored in tenant-specific Turso. The Segue Developer API receives music search criteria and track IDs and provides temporary audio URLs. The app fetches store schedules from Kanade and authorized announcement audio from Firebase Storage and maintains a temporary playback cache.

Push notifications

Both apps use device push tokens and routing identifiers with Google Firebase Cloud Messaging, via Apple Push Notification service on iOS. Play uses silent alerts for schedule changes and session revocation. Staff uses alerts for messages and urgent calls. Push payloads omit post text and recordings.

Staff posts and recordings

We use display names, text, recordings, recipients, timestamps, and urgency to deliver store communications and manage safety. Text and delivery records are stored in Google Cloud Firestore and recordings in Firebase Storage. The microphone is used when you choose to record and grant permission. Text and recordings subject to safety checks are sent to Google Gemini API. Published messages are available only to authorized recipients at the same store. Unpublished content is available to the sender and authorized reviewers.

Staff speech generation

When you choose to create a store announcement, Google Gemini API receives the input text and the selected voice setting to synthesize speech. The resulting audio is saved in Firebase Storage for reuse and delivered to authorized devices at the same store. Announcement settings are stored in tenant-specific Turso, and credit usage in shared Turso.

Consent, reports, and case records

We keep the Staff consent version and timestamp, reports, blocks, and investigation actions for permissions, delivery controls, and safety. Tenant-specific Turso holds the records, Firestore holds the necessary authorization data, and evidence audio is restricted to designated reviewers in Firebase Storage. If you contact support, we use your contact details and request to verify your identity and respond.

Retention and deletion

We keep data as needed for authentication, playback, delivery, safety, and support. Sign-out or revocation clears local credentials and temporary files and disables the corresponding push registration. After identity verification and confirmation of a deletion, access is stopped and ordinary data in scope is removed within 30 days. This includes account and device records, identifiable playback history, and affected posts and audio. Necessary safety-case evidence is removed within 90 days after case closure. Legally required exceptions are limited by purpose and time, and the records are removed after the hold ends.

External services and your choices

External services and store recipients receive the data needed for the purposes above. Tenant databases use the relevant region in Tokyo, Ireland, or the United States. The shared database and audio storage are hosted in the United States. See Google’s Gemini API terms for its handling conditions. You can change notification and microphone permissions in device settings. Contact support to withdraw consent for third-party AI processing or to request access, correction, or deletion.

Gemini API Additional Terms · Delete your account and data