Skip to content

preetharaj/AuditML-DocsAI

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

2 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Regulatory-AI-Governance

Governed NLP system for regulatory document classification with audit-ready, policy-enforced AI.

This project demonstrates how to combine ML, governance, and auditability in a practical regulatory AI system.

# Regulatory AI Classification System

**Description:**

This project implements a **governed NLP system** for classifying regulatory documents. It combines interpretable machine learning with **policy-enforced, audit-ready AI**, ensuring safe and compliant use in regulated environments.


## Key Features

- **Synthetic & versioned datasets** for reproducible experiments

- **Interpretable NLP classification model** (TF-IDF + Logistic Regression)

- **Explainability & risk analysis** for each prediction

- **Governance & model card**: links dataset versions to model versions, defines approved use cases, and enforces risk controls

- **Policy-enforced inference**: automatically blocks prohibited use, triggers human review for low-confidence predictions, and logs all actions

- **Audit-ready logging**: every decision is recorded for regulatory traceability

- **Executive demo UI**: simple, non-technical interface to demonstrate safe AI usage


## Project Structure

Reg-pj/

├── data/ # Synthetic dataset

├── models/ # Trained ML model

├── governance/ # Model card, approvals, policies

├── logs/ # Inference audit logs

├── src/ # Python source code (data prep, training, inference)

└── demo/ # Streamlit demo UI for executive presentation


## How to Run

1. **Create and activate a virtual environment**

From Command Prompt:

python -m venv venv

venv\Scripts\activate # Windows

streamlit run demo/app.py

About

Governed NLP system for regulatory document classification with audit-ready, policy-enforced AI.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages