metadata
title: Document Redaction Agent
emoji: 📄
colorFrom: blue
colorTo: red
sdk: docker
pinned: false
short_description: A tool to redact PDFs via your own AWS account
Document Redaction Agent
A two-phase PDF redaction tool powered by AWS Textract and AWS Comprehend Medical.
Phase 1 — Review: Upload a PDF → detects PHI entities → generates a highlighted review PDF and editable redactions JSON.
Phase 2 — Redact: Edit the JSON to exclude false positives, then apply permanent redactions and download the final PDF.
Authentication
Log in via IAM Identity Center (SSO) or Static Credentials. The authenticated user needs S3, Textract, and Comprehend Medical permissions in a supported region.
Requirements
- An AWS account with Textract and Comprehend Medical access
- An S3 bucket in the same region to stage PDFs for Textract processing