Product Documentation

Learn how PDF Redaction handles uploads, detects sensitive data, and produces clean files without sending document contents to a server.

Quick Start

Start Here

Open the dashboard or playground in your browser

Privacy Model

Document processing stays local to your device

Local Processing

Analysis and redaction happen in-browser without transmitting document contents

Reviewable Workflow

Inspect detected entities and download a redacted file with black box overlays applied locally

How the Product Works

1Upload a document

Choose a PDF or supported text file from the dashboard or playground. The file stays on your machine during processing.

What happens next

Browser extractionautomaticText and layout data are read locally in the browser

Supported inputs

formats.txt

PDF
TXT
CSV
JSON
XML
HTML
Markdown

2Detect PII

The analyzer finds names, emails, phone numbers, government IDs, financial identifiers, and more directly on-device.

Detection controls

Entity selectionoptionalFocus analysis on the entity types you care about

Confidence thresholdoptionalAdjust sensitivity to reduce or expand matches

Allow listoptionalExclude approved terms from being marked as sensitive

Entity coverage

entities.txt

PERSON
EMAIL_ADDRESS
PHONE_NUMBER
US_SSN
CREDIT_CARD
IBAN
LOCATION
ORGANIZATION
DATE_TIME

Example results

json

{
  "results": [
    {
      "entityType": "PERSON",
      "start": 8,
      "end": 18,
      "score": 0.85
    },
    {
      "entityType": "EMAIL_ADDRESS",
      "start": 22,
      "end": 37,
      "score": 1.0
    },
    {
      "entityType": "PHONE_NUMBER",
      "start": 46,
      "end": 58,
      "score": 0.95
    }
  ]
}

3Apply redaction

Detected entities are permanently removed from the PDF and replaced with solid black rectangles, preserving the original layout.

How redaction works

Text removal

Sensitive bytes are erased from the PDF content stream — not just covered

Black overlay

A solid black rectangle is drawn over the redacted area so nothing is visible

Layout preserved

The rest of the document — formatting, images, tables — remains intact

Common workflow

workflow.txt

1. Detect sensitive values
2. Review matches in the side panel
3. Confirm redaction targets
4. Re-run if needed
5. Download the clean document

Resources

Dashboard

Available

Use the full document workflow for uploads, review, redaction, and local downloads.

Playground

Available

Test detection rules quickly against sample or real files before broader rollout.

Industry Guides

Available

Review industry-specific pages for legal, healthcare, government, and financial use cases.

Privacy Checks

The easiest way to validate the product is to inspect it while you use it. The privacy guarantees are observable in the browser.

1Upload a document — AI detection runs locally in your browser, no data is sent to our servers

2Review highlighted entities before applying a final redaction pass

3Add custom selection areas for additional redaction (optional)

4Download the cleaned document and visually verify the output

Plans

10/day

Free Tier (live now)

Soon

Professional

Soon

Enterprise

The Free tier is currently the only enabled plan and includes 10 documents per day for registered users. Professional and Enterprise are planned next and will keep document processing local.

Ready to Get Started?

Try the product in the interactive playground or create an account for free-plan access.

Try Playground View Pricing