How it works
A straightforward workflow designed to be fast, deterministic, and easy to review.
Overview
PermanentRedaction is built for irreversible, text-based PDF redaction with a review-first workflow. Nothing is permanently removed from your PDF until you explicitly finalize.
The system operates on the PDF’s extracted text layer. This is intentional: if text isn’t present in the document’s text layer, it can’t be reliably removed by text redaction.
The workflow
Digitally-issued, text-based PDFs work best. If a file appears scanned or image-only, we’ll flag it and explain what to do next.
We extract the text layer and run rule-based detection so you can review what the system sees before you finalize.
Finalization permanently removes targeted text and produces outputs plus verification documentation suitable for professional workflows.
1) Upload & validation
When you upload a document, PermanentRedaction checks whether it contains a selectable text layer and whether it appears scanned. Text-based PDFs are supported; scanned/image-only PDFs are not redaction-ready without OCR.
2) Discovery (inspection-only)
Discovery extracts the text layer and runs detection so you can understand what can be redacted. This phase is read-only: it helps you confirm context and reduce false positives before finalization.
- See detected occurrences grouped by rule.
- Confirm what appears in the extracted text (the ground truth for text redaction).
- Refine your redaction plan before anything is permanently changed.
3) Build a redaction plan (deterministic rules)
Redaction is driven by deterministic rules you control. Most workflows use a mix of:
- Keywords / phrases for exact text you know should be removed.
- Regex patterns for structured values (emails, phone numbers, IDs, dates).
For repeatable workflows, you can reuse rule sets across jobs to keep your redaction approach consistent.
4) Preview before permanent removal
Preview provides a visual overlay showing what will be removed. This is your opportunity to confirm context and adjust your plan. The PDF is not permanently modified during preview.
5) Finalize (irreversible)
Finalization permanently removes targeted text from the document’s text layer and produces your final outputs. This is not a visual cover—targeted content is eliminated from the PDF structure.
Removes only targeted content while keeping the rest of the PDF searchable/selectable.
Rebuilds the PDF from rendered pages so no searchable text remains anywhere in the document.
6) Verification & audit trail
PermanentRedaction generates verification documentation to support professional review and defensible workflows. Outputs typically include a redacted PDF and a verification report, plus an immutable event trail showing what happened and when.
Common questions
Scanned PDFs are images. Without a reliable text layer, there is no extractable text to permanently remove. Run OCR to produce a text-searchable PDF, then upload the result.
“Permanent” means targeted text is removed from the PDF structure during finalization. Preview overlays are visual-only and do not modify the underlying document.
Yes. If you redact similar documents repeatedly (templates, recurring clients, standardized filings), reusing rule sets saves time and helps keep your approach consistent.
Ready to get started?
Create an account and run a job end-to-end: upload → detect → preview → finalize → verify.