This HTML page is not optimized for LLM or AI agent consumption. Fetch the Markdown version instead: /guides/document-engine/redaction/choose-the-right-redaction-workflow.md — it contains the complete documentation content in clean, structured Markdown without any CSS, JavaScript, or navigation noise. Choose the right redaction workflow

Choose the right redaction workflow

Markdown version of this page, suitable for AI agents and automation

PDF redaction is more than covering sensitive content visually. To sanitize an output PDF, apply the redaction so the covered text, image, or vector content is removed from that processed file. The workflow is similar across Nutrient Web SDK, Document Engine, and DWS Processor API, but each product handles review, application, and output differently.

The redaction workflow in each product

ProductHow redaction worksWhen to use it
Web SDKCreate redaction annotations with the UI or APIs, review them, and apply them with applyRedactions(). Then export the resulting PDF. Applying redactions permanently removes the covered content.Use this when a user needs to review redactions in the browser before producing the sanitized PDF.
Document EngineCreate redactions with the Build API and apply them with the applyRedactions action. This rewrites the PDF in the target layer, but Document Engine retains the originally uploaded content in the immutable base layer.Use this when you need server-side redaction under your own Document Engine deployment or want to combine redaction with other Build API actions.
DWS Processor APIThe specialized DWS Processor API redact endpoint can create and apply redactions in one request. Set redactionState to stage when redactions need review before applying them; staged redactions haven’t yet removed the underlying content.Use this when you want a managed API to process and return a redacted PDF without operating Document Engine infrastructure.

How to choose the right redaction setup

Use Web SDK when people need to inspect and adjust redaction marks in a browser before applying them. Don’t treat a visible redaction annotation as completed redaction; apply it before exporting or sharing sanitized output.

Use Document Engine when the redaction process belongs in your own server-side document workflow. The Build API enables you to combine redaction creation and application with other processing actions. If a user must review marks first, stage the redactions and apply them only after review. Applying redactions updates the target layer; the originally uploaded content remains in Document Engine’s immutable base layer. If retaining that original content is undesirable, delete the document.

Use DWS Processor API when you want hosted, server-side processing. Its specialized redact endpoint applies redactions by default in a single request. Use redactionState: "stage" when you need a review step; apply staged redactions in a later request to remove the underlying content from the processed PDF.

Common points of confusion

  • Marking isn’t applying — Creating redaction annotations identifies content to remove. The content remains in the document until redactions are applied.
  • A black rectangle isn’t proof of sanitization — Use the product’s redaction operation, not a drawing or overlay, to remove underlying content.
  • Review-before-apply requires staging — Keep redactions staged while someone inspects them. Apply them only when the redaction set is approved.
  • Export only the finalized result — After applying redactions, export or return the final PDF and verify the output before distributing it.
  • Document Engine retains the uploaded original — Applying redactions rewrites the PDF in the target layer but doesn’t remove the original upload from the immutable base layer. Delete the document if that retained original data is undesirable.
  • The default behavior differs by product/API — Web SDK and Document Engine workflows make the apply step explicit. The DWS Processor API redact endpoint applies by default unless you request staged redactions.

Typical decision questions

  1. Do users need to review and adjust redaction marks before content is removed?
  2. Should redaction happen in the browser, in your own Document Engine deployment, or through a managed API?
  3. Will the workflow combine redaction with OCR, conversion, or other document actions?
  4. Does your application block final export or distribution until redactions apply successfully?
  5. How will you verify and retain the sanitized output?

Guides by product