Send us a document. We’ll tell you where it belongs.
Every file you submit is inspected on arrival — what kind of document it is, what it contains, how long it must be kept — and returned with a suggested structure for your personal file. Nothing is shared; nothing is sold.
- Any format: PDF, scans, photos, Word, spreadsheets.
- One record per person, built file by file.
- You get the proposed folder tree and file name, not just a “thanks”.
Submit a document
Add one or more files, say who they belong to, then submit. You can correct the owner, category or retention period on the confirmation screen.
How the inspection works
- Identify. The file name, MIME type and the text pulled from the document are matched against a ruleset of personal-record document types — identity papers, tax filings, employment records, medical notes, insurance, property, education, legal, financial.
- Extract. The detector pulls the dates, identifiers and issuing authority it recognises, and flags anything it cannot read (an image-only scan, a password-protected file).
- Score. Each candidate category is scored on keyword matches, identifier patterns and date signals. A winner above the confidence threshold becomes the suggested category; anything below it is sent to manual triage rather than guessed at.
- Structure. The file is given a canonical name and mapped into a folder tree — person → life domain → document type → year — with a stated retention period and a suggested backup tier.
Handling
Encrypted at rest
The raw file lands in object storage with server-side encryption; only the extracted metadata is indexed for search.
Access is logged
Every read of a stored document is written to an append-only audit trail with the actor, the time and the reason.
Deletion is real
A deletion request removes the object and the index entry, and records the removal itself as the only remaining evidence.