Compliance-Sentinel Pro ⚖️

An Agentic AI Legal Auditor for Local, Private, and Secure Contract Analysis

Compliance-Sentinel Pro is a privacy-first, agentic AI system designed to automate the auditing of legal documents on local hardware only. Built using an Agentic RAG (Retrieval-Augmented Generation) architecture, this project goes beyond basic "chat-with-PDF" tools by actively reasoning over legal clauses to detect compliance risks, regulatory red flags, and contract gaps.

Unlike cloud-based AI platforms, Compliance-Sentinel Pro ensures 100% data sovereignty. All document processing and inference runs locally via Ollama (Llama 3.2), meaning sensitive contracts never leave the user's machine and no API keys or external services are required.

Key Highlights

• Privacy-first architecture with fully local execution
• Agentic legal reasoning using LangChain agents
• Session-isolated vector storage with ChromaDB
• Optimized performance with low-latency retrieval
• Automated professional PDF audit reports
• Secure by design for confidential legal workflows

Technical Stack

• LLM Engine: Ollama (Llama 3.2)
• Backend: FastAPI (Python)
• Frontend: Streamlit
• Agent Orchestration: LangChain
• Vector Database: ChromaDB
• Document Parsing: PyPDF2 / LangChain loaders
• Reporting: FPDF

Limitations

• Requires at least 8GB RAM (16GB recommended)
• Not a replacement for professional legal advice
• Complex legal language may require human review
• Scanned PDFs require external OCR preprocessing

Getting Started

Install Ollama and pull the model:

ollama pull llama3.2

Clone the repository and install dependencies:

git clone https://github.com/MehulMathodia/Compliance-Sentinel-Pro.git
cd Compliance-Sentinel-Pro
pip install -r requirements.txt

Run the application:

uvicorn backend.main:app --host 0.0.0.0 --port 8000
streamlit run frontend/app.py

Disclaimer

This project is intended for educational and assistive purposes only and does not replace professional legal counsel.