Skip to main content
flexorch-audit is an open-source library that detects PII, masks sensitive text, and scores document quality — entirely on your own machine. It makes no API calls and requires no FlexOrch account, so your data never leaves your infrastructure.

​
Features

  • PII detection — 46 types across TR, EU, and US jurisdictions
  • Text masking — 4 strategies: redact, replace, token, and hash
  • Quality metrics — grade (A–D), numeric score, noise ratio, and language detection
  • Compliance summary — KVKK / GDPR risk categories via compliance_report()
  • LLM preparation — redact_for_llm() one-liner that detects and masks in a single call
  • LangChain & LlamaIndex — drop-in loaders with built-in quality filtering
  • Zero dependencies — pure Python / pure JavaScript, no heavy ML models required

​
Quick example

from flexorch_audit import audit

result = audit("""
  Invoice from Acme Ltd.
  Contact: john.doe@acme.com
  IBAN: TR33 0006 1005 1978 6457 8413 26
  Amount: €12,500
""")

print(result["quality_grade"])        # "A"
print(result["pii_summary"]["count"]) # 2
print(result["detected_language"])    # "en"

​
Relationship to the FlexOrch platform

flexorch-audit is the core detection engine that powers the FlexOrch platform. When you use the platform, you get everything in this library plus:
  • Structured extraction for 9 document types
  • Managed async pipeline with job history
  • Dataset building and export
  • Cloud connectors (S3, GCS, Azure Blob)
  • Full GDPR Article 30 ROPA and NIS2 audit export
  • Team management and access controls
Use flexorch-audit when you want detection inside your own pipeline. Use the FlexOrch platform when you want the full managed workflow without building it yourself.

​
Packages

flexorch-audit on PyPI

Python 3.10+ · pip install flexorch-audit

@flexorch/audit on npm

Node 18+ · npm install @flexorch/audit

​
Explore the docs

Installation

Install the library and run your first audit in under two minutes.

Masking

Replace PII with redacted labels, tokens, synthetic values, or hashes.

Compliance Report

Generate a local KVKK / GDPR risk summary from detected findings.

LangChain Integration

Load quality-filtered, PII-masked documents into your LangChain chain.

LlamaIndex Integration

Feed privacy-safe documents into a LlamaIndex vector index.