Core privacy position
Sipura is designed around a local-first archive model. Private story material remains on-device by default unless you explicitly use a feature that requires bounded cloud processing, such as AI assistance, encrypted backup, collaboration, or web publishing.
Information we collect
- Account information such as email address, authentication identifiers, and entitlement status.
- Content you create in Sipura, including memories, notes, transcripts, story structures, and public pages you choose to publish.
- Audio recordings and microphone input when you intentionally record voice content.
- Technical information needed to operate and secure the service, such as device metadata, logs, diagnostics, and abuse-prevention signals.
- Billing and subscription information through the applicable platform and entitlement provider.
How we use information
- Provide recording, AI processing, archive organization, backup, collaboration, and publishing features.
- Generate structured outputs and summaries from the content you choose to process.
- Maintain security, prevent abuse, investigate incidents, and troubleshoot service issues.
- Manage subscriptions, usage limits, and product access controls.
What publishing makes public
When you publish a story, Sipura creates a public-safe export snapshot. That snapshot may include the approved title, selected narrative, selected thematic context, and public-safe media. The private archive is not mirrored wholesale to the web.
Unlisted pages are intended to stay out of public discovery and search indexing, but anyone with the direct URL may still access them. Unlisted is not the same as private or access-controlled.
What stays private by default
- Raw audio files unless you explicitly share or back them up.
- Full transcripts and rough notes unless you explicitly include them in a public-safe export.
- People, place, and theme relationships that exist only inside the private archive.
- Full version history and internal processing context.
Third-party providers
- Supabase for backend data and authentication infrastructure.
- OpenAI for AI-assisted processing where applicable.
- RevenueCat for subscription entitlement handling where billing is enabled.
- Sentry for crash and stability monitoring when enabled.
Retention and deletion
We retain information for as long as needed to operate the service, meet legal obligations, resolve disputes, and enforce agreements. Public story pages remain online until removed, archived, or replaced through the publishing controls.
To request deletion of account-linked cloud data or to ask privacy questions, contact contact@glasrocks.com.
International handling and legal rights
Because Sipura uses cloud infrastructure providers, data may be processed in jurisdictions outside your home country where permitted by law and necessary to operate the service.
Depending on your jurisdiction, you may have rights to access, correct, delete, or object to certain processing. Contact us to make such a request.
Contact
Privacy requests and questions can be sent to contact@glasrocks.com.