RBI Notifications RAG System

Overview

This project implements a Retrieval-Augmented Generation (RAG) system specifically designed to query RBI (Reserve Bank of India) notifications. It leverages the LLaMA index and FastAPI to create a responsive and informative query system. The primary goal is to provide users with specific information contained within RBI notifications through an easy-to-use interface.

Installation

Prerequisites

• Python 3.8+
• pip

Steps

1. Clone the repository: git clone <repository-url>
2. Navigate to the project directory: cd <project-directory>
3. Install the required Python packages: pip install -r requirements.txt
4. Create .env file and store openAI API key: OPENAI_API_KEY= Enter your api key here

Usage

To start the FastAPI server and access the web interface: 'uvicorn fast_api:app --reload' Navigate to http://127.0.0.1:8000/ in your web browser to interact with the system.

Code Explanation

• main.py: Initializes the query system, setting up the context, available tools, and the ReActAgent for handling queries.
• create_index.py: Handles the creation and loading of the LLaMA index used for querying the RBI notifications dataset.
• pandas_query_engine.py: Sets up a PandasQueryEngine for querying the notifications dataset directly using pandas.
• save_notes.py: Provides functionality to save text-based notes to a file, allowing users to store information retrieved from the system.
• fast_api.py: Contains the FastAPI app setup, including routes for displaying the form and handling queries.
• form_page.html: HTML template for the web interface, allowing users to submit queries and view responses.

Packages Used

• dotenv: Loads environment variables from a .env file, used here to manage API keys securely.
• llama_index: A library for building and interacting with vectorized search indexes, used to create a powerful query engine for RBI notifications.
• FastAPI: A modern, fast (high-performance) web framework for building APIs with Python 3.7+ based on standard Python type hints.
• pandas: Provides high-performance, easy-to-use data structures, and data analysis tools for Python.
• Jinja2: A full-featured template engine for Python, used here for rendering the web interface.

Outputs

See all outputs in the output_images folder

Future Improvements

• Expand the dataset to include more comprehensive RBI notifications.
• Enhance HTML parsing capabilities to extract and utilize more structured data from notifications.
• Evaluate different large language models for potential cost reductions and performance improvements.
• Integrate additional query engine tools to expand the system's capabilities.
• Improve the user interface to include features like question and response history.

Contact

For any queries or contributions, please email: [email protected].