Features
True PDF redaction.
Not a black box over sensitive data.
Most tools draw rectangles on top of your document and call it redaction. True PDF redaction destroys the sensitive data itself — pixels, text layer, metadata, and all. That difference is the whole product.
Visual masking vs. true redaction
The critical difference is what happens to the data underneath.
| Visual maskingtypical PDF editors | True redactionDr. Redact | |
|---|---|---|
| What the 'redaction' is | A black rectangle drawn over live text | Pixel-level destruction burned into a rebuilt file |
| Copy & paste the 'hidden' text | Works — the text is still there underneath | Impossible — no text layer exists in the output |
| Metadata, comments, revision history | Travel with the file | Never carried over — the output is built from scratch |
| How failures are discovered | By the recipient (or the press) | By you, in the built-in verification step |
The 3 most common redaction mistakes
1. Drawing a box over the text
In most PDF editors a black rectangle is an annotation sitting on top of live text. Select-all, copy, paste — everything 'redacted' appears in plain text. Courts, agencies, and journalists have exposed exactly this failure again and again.
Dr. Redact never overlays. Approved redactions are burned into the page pixels and the original text layer is discarded entirely.
2. Trusting the text layer of a scan
A phone photo, fax, or scanner PDF has no reliable text layer. Tools that only search text will report 'nothing found' while the SSN sits right there in the image — or miss handwritten entries completely.
Dr. Redact detects image-based pages and runs OCR automatically, including tilted phone photos and handwriting, plus dedicated signature detection.
3. Forgetting what you can't see
PDF metadata, author names, comments, tracked changes, embedded revision history, attachments — hidden content ships with the file and can expose people and organizations even when every visible word was handled.
Dr. Redact generates a brand-new file: every page carrying a redaction becomes a flattened image, untouched pages are carried over with comments and annotations stripped, and metadata is blanked. Our verification suite plants a fake reviewer comment and confirms it is destroyed.
Scanned documents
Why scanned PDFs require OCR to redact properly
A scanned page is a photograph. There is no text for a text-search tool to find — so it finds nothing, reports success, and ships the SSN untouched. Real-world documents are tilted phone photos, faxes, and photocopies, and they need to be read the way a human reads them.
Automatic OCR fallback
Pages without a usable text layer are read with enterprise OCR — no settings to remember.
Phone photos & tilted scans
Skewed camera shots and low-quality copies are handled, including handwritten entries.
Scan straight from your phone
No scanner needed: on the upload page, snap paper pages with your camera — each photo becomes a page of one PDF, built on your device. Dr. Redact never asks for camera access; you hand it exactly the pages you choose.
Signature detection
Handwritten signatures are found visually and offered for redaction like any other match.
High Sensitivity mode
For documents where missing one item is unacceptable: every category, expanded AI context, lower-confidence matches surfaced for review.
Hidden data
Metadata, comments, and revision history: sanitized by construction
Dr. Redact doesn't scrub your file — it rebuilds it. Every page carrying a redaction becomes a flattened image (with an invisible, searchable text layer laid back over it), and untouched pages are carried over with their annotations stripped. Original metadata, author names, comments, tracked changes, and revision history simply do not exist in the output, so they can't expose your organization — and the file's own timestamps are zeroed.
The legal & financial stakes
A failed redaction isn't an embarrassing typo — it's a data breach. Under HIPAA, exposed PHI carries civil penalties that scale per violation with annual caps in the millions, plus breach-notification duties. In litigation, botched redactions have produced court sanctions, sealed material in the public record, and front-page news. For any regulated business, the cost of one leaked document dwarfs the cost of redacting ten thousand properly.
That's why Dr. Redact adds a step no black-box tool has: after processing, you review the actual burned output before you download it — and if anything else needs to go, additional rounds on the same document are free.
65+ detection categories, reviewed by you
Pattern precision plus AI context — and every single match is shown for your approval before anything is permanent.
Security architecture
Encrypted with a dedicated key
AES-256 encryption at rest under a dedicated KMS key, TLS in transit, and time-limited access links that expire in minutes.
Deletion that means deletion
Files auto-expire on your plan's retention schedule, and manual deletes are real storage deletions — every one recorded in a permanent audit log.
Compliance paperwork built in
Audit certificates on paid plans document every redaction. BAA execution is click-wrap, timestamped, and stored immutably.
Full details on the compliance page.
Frequently asked questions
Is the redacted text really gone, or just hidden?
Gone. Every page carrying a redaction is rasterized — the bars are burned into the pixels and the original page content is destroyed, not covered. The redacted values exist nowhere in the output file: not in a text layer, not in the file's raw bytes. We verify this mechanically on every engine change by searching the output bytes for the redacted strings.
Is the redacted PDF still searchable?
Yes. Pages without redactions keep their original text untouched. Pages with redactions are rebuilt as images for safety, then every word that touched no redaction box is laid back over the image as an invisible text layer — so search, select, and copy-paste work across the whole document, while the redacted words themselves were never written into the file in any form.
Does it remove PDF metadata, comments, and revision history?
Yes — and we prove it rather than assume it. The output file is generated from scratch: author info, title, comments, annotations, and revision history are stripped, and even the file's own timestamps are zeroed so it doesn't reveal when it was processed. Our verification suite plants a fake reviewer comment in a test document and confirms it is destroyed.
Can it handle scanned documents and phone photos?
Yes — and they come back better than they went in. Pages without a usable text layer are read with OCR, including tilted phone-camera shots and handwriting, and signatures are detected visually. You review every match before anything is burned. Then, at processing, scanned pages are OCR'd again from the redacted image and given an invisible text layer — so your scan comes back as a searchable document, with the redacted content physically unreadable to the recognition step.
Can I scan paper documents with my phone camera?
Yes. On the upload page, tap "Scan with your camera" — snap each page and they become one PDF, built entirely on your phone before upload. Dr. Redact never requests camera or photo-library permission: your phone's own camera app takes the photos and hands over only the shots you choose. Crooked pages are fine — OCR reads them — and the redacted result comes back searchable like any other document.
What if the AI misses something?
You review every detection before processing, can draw manual boxes, and can rescan with more categories without losing your work. After processing you verify the actual output — and additional redaction rounds on the same document are free.
Do you keep a copy of my original document?
Only until you process it. Up to that point the original is stored encrypted so you can scan, review and rescan as many times as you like. The moment you press Process Document, the redacted output replaces it and the unredacted original is destroyed — your audit certificate states this. Afterwards you can still add more redactions for free, because anything the scan missed is by definition still visible in the redacted file. What cannot be undone is a redaction you decide you didn't want, so review before processing.
Is Dr. Redact HIPAA compliant?
Dr. Redact is built for HIPAA workflows: encrypted processing, retention controls, audit certificates, and a Business Associate Agreement executed click-wrap on paid plans. The free trial does not permit PHI.