Back to services
API OverviewDocument Intelligence

PII Detection & Redaction API

Detects supported categories of personal information and can return redacted document content according to the service contract.

When to use this API

PII Detection & Redaction API supports GDPR/RODO-oriented workflows where sensitive identifiers must be removed before sharing. APIRAPTOR performs extraction, deterministic detection, optional contextual AI detection, controlled redaction and post-redaction validation before returning the final artifact. The service supports compliance workflows but is not a standalone legal guarantee.

  • You need to anonymize contracts, support exports or HR files before sharing.
  • You need auditable findings without exposing raw PII by default.
  • You need one API flow that supports redacted file output, package output and detect-only analysis.

Typical use cases

  • GDPR/RODO compliance redaction before handing documents to support teams.
  • Cleaning personal data from logs and ticket exports.
  • Redacting CVs, agreements and sales documents for controlled sharing.

Billing and APIRAPTOR token model

PII Detection & Redaction usage is metered per processed page through the `processed_page` service billing event.

  • Tokens are the APIRAPTOR communication unit: 1 token = 100 subtokens.
  • Trial availability, token amount and duration come from Billing configuration.
  • All services are available from day one; PII Detection & Redaction is not premium-gated per plan.
  • This is service billing event metering, not ordinary API-call billing.

Redaction quality boundary

PII Detection & Redaction supports privacy workflows but does not replace legal review or guarantee complete regulatory compliance.

  • Detection quality depends on file readability, language, context, enabled categories and policy configuration.
  • AI findings below configured confidence are review signals and should not be treated as automatic final decisions.
  • The service does not guarantee every personal-data occurrence will be detected or redacted without human review.
  • Detect-only output is a review artifact; file/package modes are the redaction output paths.

Sync and async usage

Sync and async requests use the same canonical PII Redaction POST endpoint.

  • Sync uses POST /v1/pii-redaction with mode=sync or the configured default mode.
  • Async create uses POST /v1/pii-redaction with mode=async.
  • Jobs endpoints are used for status/result metadata, not job creation. File artifacts are returned as file_id values for authenticated API download or signed URLs when direct download is enabled, and generated file artifacts are deleted automatically no later than 48 hours after creation.

Input data

Upload source file (PDF/DOCX/JPG/JPEG/PNG/WEBP) or provide inputText, then choose profile, categories, replacement style, output mode and output format.

Output data

Default response is a redacted binary file. package mode returns ZIP (file + findings report). detect_only mode returns JSON findings.

How it works

  1. Input is validated, scanned and stored using APIRAPTOR artifact storage.
  2. Deterministic validators (checksum/regex) run first, then optional contextual AI findings are normalized into a controlled internal model with confidence thresholds.
  3. APIRAPTOR applies redaction, validates residual exposure risk and returns file/ZIP or detect-only findings based on outputMode.

API documentation

Review endpoint parameters, request/response examples, error codes and execution notes in the technical API reference.

View API docs

Related APIs

Start using this service

Continue with endpoint details in documentation or open your workspace in the dashboard to run live requests.

Syncing service registry...