Legal PDF redactor — find sensitive data locally

Nothing uploadedTested up to 256 MB

Scan a PDF locally for common sensitive identifiers and custom patterns, review every match, then permanently redact approved regions without uploading the document.

Review a legal-production PDF for common sensitive identifiers before disclosure. The browser-local scanner can find Social Security numbers, EINs, U.S. phone numbers, email addresses, payment-card numbers that pass a Luhn check, dates of birth, and custom regular expressions. Every proposed rectangle stays editable before the existing destructive redaction engine rebuilds affected pages and performs its privacy verification checks. Pattern matching only examines extractable PDF text: scanned pages, unusual text encoding, fragmented text runs, names, addresses, privilege content, and context-specific confidential information can require manual review. The PDF is processed locally in your browser and is not uploaded to a document-processing server, up to 256 MB.

Processed locally in your browser

PDF sanitization workspace

The PDF stays on this device. Select regions only when you accept rasterization of those pages.

PDF and redaction regions

Drop a PDF hereor choose a filePDF up to 256 MB · processed locally

Raster output

Raster resolution (DPI)

150 DPI is the default. Higher values improve small text but increase processing time and file size.

Marked pages will become images

This is the destructive step that makes region redaction real. Every marked page loses selectable text, searchability, accessibility structure, and usually becomes larger. Unmarked pages keep their vector content and text layer.

Hidden-data removal

Strip document metadata and XMPRemoves title, author, subject, keywords, producer, creator, dates, and the XMP metadata stream.Keep creation and modification datesPreserves only the two PDF timestamp fields while other selected metadata is removed.Remove embedded attachmentsEmbedded files can contain complete hidden documents. Leave this enabled unless they are required.Remove annotations and form widgetsComments, links, widgets, and their hidden content are removed rather than presented as flattened proof.

The output is a rebuilt PDF

Rebuilding prevents removed page objects from surviving as unused data, but signatures are invalidated and document-level features such as bookmarks, scripts, forms, and structure tags may not survive. A source with incremental revisions is reported separately; this tool does not certify removal of historical revisions.

Review a legal-production PDF for sensitive identifiers before disclosure

1
Choose the production PDF and let the browser render local previews.
2
Enable the sensitive-data patterns that apply to the matter and add any matter-specific regular expressions.
3
Scan the extractable text, then inspect every proposed redaction rectangle.
4
Adjust, add, or remove regions manually and choose hidden-data cleanup options.
5
Create the rebuilt PDF, review the verification report, and perform your normal production QA before disclosure.

Built for the real task

Six built-in sensitive-data patterns
Custom regular-expression patterns
Reviewable and manually adjustable detection rectangles
Destructive redaction with hidden-data cleanup

See the transformation

Hidden PDF data is removed, and marked pages become final pixels.

Remove hidden document details and permanently cover the areas you marked.

Input

Result

Browser limits

Files are limited to 256 MB to protect browser stability, especially on phones. Processing stays on this device.

Six built-in sensitive-data patterns

Scan a PDF locally for common sensitive identifiers and custom patterns, review every match, then permanently redact approved regions without uploading the document.

Private by design

The selected file is not uploaded to Nablyx.

Processed locally in your browser

No. This tool reads and transforms the file locally in your browser.

Common uses

Six built-in sensitive-data patterns
Custom regular-expression patterns
Reviewable and manually adjustable detection rectangles
Destructive redaction with hidden-data cleanup

Does automatic pattern detection guarantee that a legal production is fully redacted?

No. The scanner only matches configured patterns in extractable PDF text. Scans without a text layer, fragmented or unusually encoded text, names, addresses, privileged material, and context-specific confidential information can be missed. Treat detection as a review aid and perform the legal and production QA required for your matter.

Legal Production PDF Redaction Checklist Before Disclosure

What Pattern-Based PDF Redaction Can Miss

Reviewing Court Privacy Fields Before Filing or Production