Simpli Redact

Redact makes your support data safe for AI consumption. Scan conversations and KB articles for credit card numbers, SSNs, emails, phone numbers, and addresses — then redact them before feeding data to AI models.

Simpli Redact provides PII detection, automatic redaction, batch risk scanning, and safety validation for support data.

Configuration

Variable	Default	Description
`APP_PORT`	`8014`	Server port
`LITELLM_MODEL`	`openai/gpt-5-mini`	LLM model for PII detection
`REDACT_REPLACEMENT`	`[REDACTED]`	Replacement string for detected PII
`DETECT_TYPES`	`credit_card,ssn,email,phone,address,name,account_number`	PII types to detect
`CORS_ORIGINS`	`*`	Allowed CORS origins (comma-separated)

Start the server

simpli-redact serve

API endpoints

All endpoints are under the /api/v1 prefix.

POST `/api/v1/redact`

Redact PII from text, returning sanitised versions.

Request:

{
  "texts": [
    {"text": "My card is 4111-1111-1111-1111 and email is john@example.com"},
    {"text": "Call me at 555-123-4567"}
  ],
  "replacement": "[REMOVED]"
}

Response:

{
  "scan_id": "s-abc123",
  "total_texts": 2,
  "results": [
    {
      "index": 0,
      "original_text": "My card is 4111-1111-1111-1111 and email is john@example.com",
      "redacted_text": "My card is [REMOVED] and email is [REMOVED]",
      "redaction_count": 2
    }
  ],
  "total_redactions": 3
}

POST `/api/v1/detect`

Detect PII without redacting — returns entity locations and types.

POST `/api/v1/scan`

Batch risk summary — how much PII is in your dataset, broken down by type.

POST `/api/v1/validate`

Pre-flight safety check for a single text — returns safe: true/false.

GET `/health`

Health check.

Next steps

Model Selection — Choose the right model for PII detection

On this page