Status
PyperSafe
A personal document safe that protects digital archives and helps retrieve their paper originals.
The need
Bills, certificates, and administrative records tend to become scattered across digital folders and paper binders. PyperSafe gives them one entry point: a local desktop application that imports documents, protects their content and metadata, then retrieves the right item through structured search.
The journey has four steps: import, describe, search, retrieve. Each document receives a stable identifier. Printable labels connect that identifier to the sleeve holding the paper original.
What I designed and delivered
- A metadata catalog with search by tags, date, import batch, and sensitivity.
- A self-contained store whose location is selected by the user and can be moved.
- SQLite encryption with SQLCipher and document encryption with
XChaCha20-Poly1305. - Safe-key derivation from the passphrase using
Argon2id. - Parallel batch import with atomic publication per document and rollback when catalog registration fails.
- Image and PDF previews, decrypted export, and printable UID label sheets.
- Linux and Windows releases with reproducible illustrated documentation.
Skills demonstrated
The project combines product design, domain modelling, local security, and desktop UI work. Its hexagonal architecture separates domain concepts, use cases, and technical adapters. Development followed short TDD slices, with close attention to partial failures, catalog integrity, and long-running work outside the GUI thread.
The result is a complete, installable, documented application built around a practical workflow.