Tutorial: Detect PII Without Redacting

Build an n8n workflow that scans a PDF for names, emails, and dates using the Detect PII operation, without modifying the document.

Use the Detect PII operation when you want to know what sensitive information is in a document — for auditing, building a review step before redacting, or routing documents based on what they contain — without modifying the PDF itself.

This tutorial builds a runnable n8n workflow that fetches a public sample document and detects names, emails, and dates in it.

For the general setup/configuration walkthrough, see Setup & Configuration Tutorial first. For redacting a PDF (rather than just detecting PII in it), see Redact a PDF fetched over HTTP. This guide assumes the node is installed and you have a PDF Redaction API credential already.

1. Trigger + HTTP Request

Start with a Manual Trigger, then add an HTTP Request node with:

Method: GET
URL: https://raw.githubusercontent.com/StabRise/pdf-redaction-api/main/examples/pdfs/sample_document_en.pdf
Options → Response → Response Format: File

Execute the node — the sample document (1.11 MB) comes back as binary data in the data field:

HTTP Request node output showing the fetched sample_document_en.pdf binary data

2. PDF Redaction — Detect PII

Add a PDF Redaction node after the HTTP Request node and set:

Credential: your PDF Redaction API credential
Operation: Detect PII
Additional Fields → Tags: pick the PII types to look for — this example uses Date, Email, Person Name

Unlike Anonymize, this operation has no Output Binary Field — it doesn't produce a modified PDF, only detection results.

PDF Redaction node configured with the Detect PII operation and Date, Email, Person Name tags

3. Execute and inspect the output

Click Execute step. The output's detected_pii[].entities array lists every match, each with an entity_group (e.g. PERSON_NAME, EMAIL), the matched word, and boxes giving the pixel location on the page.

For this document, the node found entries like:

PERSON_NAME — "Johnathan Smith"
EMAIL — "[email protected]"

along with Date matches further down the list, plus the usual processing_time breakdown.

4. The complete workflow

Complete workflow: Manual Trigger → HTTP Request → PDF Redaction (Detect PII), all executed successfully

Notes

If you leave Tags empty, the API only reports the entity types it's configured to detect by default for the request — for predictable results, always set the tags you care about explicitly.
Combine Detect PII with an IF or Switch node downstream to branch a workflow based on what was found (e.g. only redact documents where detected_pii is non-empty).
To then redact what you found, feed the same input into a second PDF Redaction node set to Anonymize with matching tags — see Redact a PDF fetched over HTTP.