Compliance Monitoring and Enforcement through Log Analysis using Large Language Models

Ensuring compliance with security standards and policies becomes an intricate task with the increasing volumes of data and system complexities. This project harnesses the power of Large Language Models (LLMs) to provide insightful monitoring and robust enforcement through meticulous log analysis.

📜 Problem Overview

With companies handling vast volumes of data, the significance of maintaining security compliance is unparalleled. This initiative is specifically designed to:

• Analyze logs, system configurations, access controls, and user privileges.
• Validate compliance against security standards and policies.
• Employ LLMs, such as open-source alternatives to ChatGPT, to detect non-compliance and suggest actionable remedies.

✨ Solution Attributes

• Rule Definition: Elicit rules and deduce relationships between these rules and log/policy parameters.
• Flexibility: Process diverse log formats (CSV, text, PDF), rule sets, and compliance standards.
• Actionable Insights: Generate pertinent and actionable suggestions for remediation.
• System Performance: Handle vast volumes of log and text data efficiently.
• Adaptability: Learn and adapt from new rules, updated standards, and feedback.
• Accuracy: Ensure high precision and recall with minimal false positives and negatives.
• UI/UX (Optional): An intuitive interface for system interaction, file uploads, report viewing, and insight access.

🔄 Workflow

1. Input: Users provide compliance documents as rulesets and logs/system policies with known discrepancies.
2. Analysis: The system meticulously reviews the inputs, pinpointing compliance breaches, and providing specific citations.
3. Output: Obtain insightful recommendations to rectify the identified breaches.

🔧 Technical Requirements

• An open-source Large Language Model (LLMs) — APIs are avoided to ensure data privacy.
• A robust tech stack for fluid automation and precise parsing.

🚀 Setup and Installation

1. Clone the Repository:

   git clone https://github.com/fibonacci35813/sudo_hack.git
   cd sudo_hack/grid

2. Install Dependencies:

   pip install -r requirement.txt

3. Run the Application:

   streamlit run main.py

🤝 Contributing

Contributions are always welcome!