Mini-RAG Demo

A simple demonstration of Retrieval-Augmented Generation (RAG) that showcases the difference between vanilla AI responses and grounded, context-aware answers.

Overview

This application demonstrates a basic RAG system that:

• Loads documents from a local folder
• Retrieves relevant documents based on user queries
• Provides both vanilla AI answers and grounded answers using retrieved context
• Shows how retrieval improves answer quality and accuracy

Features

• Document Loading: Automatically loads text documents from the sample_docs/ folder
• Simple Retrieval: Uses keyword-based scoring to find relevant documents
• Dual Response Mode:
  • Vanilla answers without context
  • Grounded answers using retrieved documents
• Interactive CLI: Easy-to-use command-line interface

Installation

1. Clone or download this repository
2. Ensure you have Python 3.6+ installed
3. No additional dependencies required (uses only Python standard library)

Usage

Run the demo from the command line:

python demo.py

The application will:

1. Load all .txt files from the sample_docs/ folder
2. Prompt you to ask a question
3. Show a vanilla AI response (without context)
4. Retrieve relevant documents based on your query
5. Show a grounded response using the retrieved context

Example Session

Welcome to the Mini-RAG Demo.
Ask a question: How do I reset my password?

--- Without Retrieval (Vanilla AI) ---
I'm answering your question 'How do I reset my password?', but I don't have any documents to look at. My answer may be incomplete.

--- Retrieving Context ---

--- With Retrieval (Grounded Answer) ---
Answer based on retrieved docs for 'How do I reset my password?':

- From faq.txt (score 2): Students often ask how to reset passwords. Password reset links expire after 15 minutes for security reasons. Users must verify email ownership...

This answer is grounded in retrieved context.