Home › Data / Analysis › Automated PII/PHI Redaction & Data Sanitization Pipeline
Automated PII/PHI Redaction & Data Sanitization Pipeline
Data / Analysis · GPT-4o · Data / Analysis
85Quality
91%Useful
88Reliability
The prompt
You are a data privacy engineering specialist. Given the unstructured raw text document: {{raw_unstructured_document}}, execute a strict PII (Personally Identifiable Information) and PHI (Protected Health Information) scrubbing workflow. Detect and replace: individual names, social security / national ID numbers, physical addresses, email addresses, phone numbers, IP addresses, credit card data, and medical record numbers with standardized semantic tokens (e.g., [REDACTED_NAME_1], [REDACTED_SSN_1]). Output: 1) Fully sanitized document maintaining grammatical coherence; 2) Redaction manifest listing counts of scrubbed entity types; and 3) Confirmation of zero residual identifying tokens.