Privacy notice

Private-beta implementation draft. Operator identity, contact details and retention periods must be confirmed before new member access is enabled.

What the application stores

Signing in uses Google or GitHub. The application stores your verified email, provider identity, membership status and account preferences. Authentication uses session cookies. The player uses browser storage for listening preferences and queue/history features.

Submitted text, its source reference, generated scripts, supporting quotations, audio and job status are stored on the server. Administrators can review private drafts. Public visitors can read published summaries and play published recordings.

Where processing happens

Source text and prompts are sent to Ollama Cloud for text generation. Its handling of those requests is described in Ollama’s privacy policy. New speech synthesis runs on the application server using Kokoro; it does not require sending text to OpenAI. Railway hosts the application and persistent files. Login providers process sign-in information under their own policies.

Retention and removal

There is currently no automatic expiry for accounts, source snapshots or generated files. Server and hosting logs may contain request metadata. A retention schedule and private contact route must be established before accepting new member content.

Withdrawing an episode removes application access to its public recording. Copies already downloaded while it was public cannot be recalled. Private API responses and audio previews are marked not to be cached; the service worker stores only static application assets.

Contact

The operator’s private contact route is pending configuration. Do not put private material in public issue trackers.