This repository contains a Python-based tool that leverages Large Language Models (LLMs) and Retrieval-Augmented Generation (RAG) to assist in the analysis of Environmental, Social, and Governance (ESG) policies.

Key Features

• Identifies potential gaps and discrepancies between company policies and relevant ESG regulations.
• Utilizes OpenAI's LLM for advanced text understanding and analysis generation.
• Employs the Langchain library for flexible integration of LLMs and retrieval techniques.
• Pinecone vector database for efficient storage and retrieval of regulations.
• Streamlit integration for a user-friendly web interface.

Prerequisites

• Python 3.7 or later
• OpenAI API key
• Pinecone API key and environment
• A structured source of ESG regulations (e.g., JSON file, database)
• Required Python libraries:
  • openai
  • langchain
  • streamlit
  • spacy
  • pypdf2
  • pinecone-client

Installation

1. Clone this repository:

   git clone https://github.com/mominalix/ai-powered-esg-compliance.git

2. Install dependencies:

   cd ai-powered-esg-compliance
   pip install -r requirements.txt 

Running the Tool

1. Set your API keys in the ESG_assessment.py file or as environment variables.
2. Prepare your regulations data in the supported format.
3. Start the Streamlit application:

   streamlit run ESG_assessment.py

Usage

1. Upload your company policy documents (PDF format).
2. The tool will extract the policy text, analyze it against relevant ESG regulations, and highlight potential gaps or areas for improvement.

Disclaimer

This tool is intended for preliminary analysis and should not be considered a substitute for professional legal advice.

Future Development

• Fine-tuning the LLM on ESG-specific data.
• More sophisticated gap analysis logic using advanced NLP techniques.
• Integration with external regulatory databases.