Frequently asked questions

Is my file uploaded anywhere?

No. The document is read, processed and written back by the page itself, in your browser. There is no account and no storage. You can verify it by loading the page, disconnecting from the network, and doing the whole job offline.

Does it really delete the text, or just cover it?

It deletes it. A page carrying a redaction is turned into an image, the zones are painted onto that image, and the image replaces the page — so that page has no text layer left to recover anything from. Pages you did not touch keep their text.

What happens with a scanned document?

A scan carries no text layer, so the file itself has nothing to detect in. The page can be read by optical recognition on demand, in the browser like everything else, after which detection and word-selection work on it — but recognition guesses, so a value it misreads is never proposed, and the interface says which pages it read that way. Zones can always be drawn by hand instead, with the mouse or the keyboard, and those are painted like any other redaction.

How large a file can it handle?

Up to 100 MB and 200 pages. The ceilings exist because everything runs in your browser, whose memory is limited; a larger document has to be split and processed in parts.

Are metadata removed as well?

Yes. Document metadata, annotations, file attachments and page thumbnails are stripped from the produced file, because text removed from a page can otherwise survive in any of them.

What does striking through or underlining do?

They annotate, they do not redact. The text stays in the file and remains readable — that is what those tools are for. Only redaction removes anything, and the interface keeps the two apart on purpose.

Can it refuse to give me a file?

Yes, and that is deliberate. The produced file is re-read and searched for every string that was supposed to be gone. If any of it is still extractable, the export is cancelled and nothing is handed over, with the surviving value named so you can act on it.

Can a removed value be replaced by a label?

Yes. The « Replace with a label » tool on the action bar writes a generated label — [NAME 1], [SIN 1] — in place of each removed value, the same one everywhere that value appeared. The document stays readable: you can see that the person on page 4 is the person on page 1, which a row of black rectangles does not say. Which value a label stands for exists only for the length of the session: it is written nowhere, neither in the produced file nor in the report. A value too short to carry a label is redacted without one, and the screen says so before the export.

What is a profile, and what is in it?

A profile is a file you keep, carrying the way you handle a case: the rules you leave unticked, the terms you redact every time, patterns of your own — a case-number format, say — and whether labelling is on. Load it and all of that is applied to the open document; the last profile used comes back on its own next time. A profile unticks, it never removes: everything detected is still listed, and you can always change your mind. One thing to know: a profile holds names in clear text and it outlives the session in your browser. It is the only place in PDFSafeKit where data taken from a document stays behind — keep the file the way you would keep the document.

How much does it cost?

Nothing. There is no account, no trial and no paid tier.