Project information

Privacy Policy

Effective 2026-10-25

Mom’s Journals is a private family journal-digitization system. This policy explains how the private application handles journal material and Google account data, and how this public website differs from the archive.

Information the application uses

The application uses original journal photographs, image variants prepared for transcription, AI transcription drafts, selected human-reference examples, optional human corrections and related people, places and other journal context. It also preserves source relationships, verification evidence, processing history and usage records.

Google Drive information used for intake includes file and folder names and identifiers, parent relationships, versions, sizes, checksums, ownership, sharing permissions, account identity and capabilities. These checks help establish whether a source belongs to the authorized project and whether it may be processed.

Google Drive access

The existing Google consent connection is named “Moms Journals Local Read Only.” It supports this private journal project, not a public sign-in service on this website.

An administrator-authorized Google connection uses drive.metadata.readonly to inspect metadata and permissions. A separate Viewer-only identity uses drive.readonly to read eligible original images. The application restricts its file access to the designated Mom’s Journals project folder and its descendants; the OAuth scope itself is not a Google-enforced folder boundary.

The intake application does not edit, move or delete Google Drive files or change their sharing. Contributors upload through their existing assigned Google Drive folders. Access tokens are used for authorized processing; protected ongoing credentials are stored locally in macOS Keychain rather than in this website.

Removing the application’s Google authorization prevents future access using that grant. It does not erase originals, transcriptions or backups already preserved. Contributors should contact the administrator about their folder access or previously supplied material.

AI transcription and context

For an authorized transcription, the application sends source-derived journal images, transcription instructions and selected existing human-reference image/text examples to Google’s Gemini API. These inputs can contain personal information recorded in the journals. Google returns a draft and processing-usage information. Context and correction records are maintained in the private local database.

AI drafts may contain errors. Human review is selective and optional; confirmed human versions remain distinct from AI drafts. The project does not use Google API data to train or improve a general-purpose AI model. Sending inputs for transcription is separate from model training. Google’s processing and retention are governed by its Gemini API terms; this project does not promise zero provider retention.

Storage, access and sharing

Originals, transcriptions, context and processing evidence are stored in a private local archive and SQLite database. The backup workflow preserves consistent database snapshots and referenced files in an encrypted Restic repository on separate storage. Historical backups and additional offsite protection may retain copies.

The administrator operates the private application. Authorized family contributors use their assigned Drive folders and approved review workflows; participation does not grant access to the administrator’s computer, dashboard or the entire archive. Private journal material is not published on this website or included in the source-code checkpoint.

Google processes Drive and Gemini requests. Backup storage providers process stored backup data as applicable to the configured backup arrangement. The project does not sell journal or Google account data, use it for advertising, or send it to unrelated services for their own purposes.

Retention, corrections and deletion

The journals are maintained as a continuing family archive. The application has no automatic expiration or deletion schedule for journal records. Original evidence and transcription histories are deliberately preserved; a correction normally creates a new version rather than overwriting the earlier one.

You may request a correction, access review or deletion by contacting the administrator. There is no self-service deletion tool on this website. Removal requires a controlled review of Google Drive sources, local originals, derived records and backup history; it is not performed automatically by revoking consent. Earlier copies may remain in backups, and no fixed purge deadline is promised here.

This public website

This site has no accounts, journal upload forms, embedded journal content, advertising or analytics scripts. It does not ask visitors to connect Google Drive. Cloudflare hosts the static pages and may process connection and request information to deliver and secure them under the Cloudflare Privacy Policy. The site does not add tracking cookies or browser storage.

Google API policy

Mom’s Journals’ use and transfer of information received from Google APIs adheres to the Google API Services User Data Policy, including its Limited Use requirements. Google API data is used to provide and protect the disclosed journal-preservation and transcription functions, rather than unrelated purposes.

Contact and changes

For questions, correction requests, access reviews or deletion requests, contact the project administrator at journals@sumsion.com.

This page will be updated when the project’s data practices change. Any new Google access that requires consent must be authorized before use.