🧠 LeadRAG – Leader Embeddings with Keyword Corpus in RAG

LeadRAG is a secure and intelligent Retrieval-Augmented Generation (RAG) system that enhances information retrieval through leader embeddings and a keyword-driven corpus. It combines semantic and lexical techniques to provide high-quality, context-aware answers. Role-based access and admin capabilities ensure a controlled and customizable environment.

---

🚀 Features

• 🔍 Hybrid Retrieval: Combines leader embeddings with keyword search for improved document matching.
• 🧠 RAG-based QA: Retrieves context and uses LLMs to generate accurate answers.
• 🧱 Modular Architecture: Easy to integrate or extend with custom components.
• 🔐 Role-Based Access: Includes user and admin modes for controlled data access.
• 📊 Dashboard Views: Guided interactions and feedback visualizations

📦 Installation & Setup

1. Clone the Repository

1. Clone the Repository

git clone https://github.com/jay-3101/LeadRAG.git
cd RAG_Chatbot

2. Set Up Your GROQ API Key

This project uses GROQ’s free API to call the LLaMA model(Open Source Model.

Steps: Sign up at https://console.groq.com/keys

Generate your GROQ API key

3.Running the Project

1. Using shell Script

Use the run.sh script to automatically install dependencies, export your API key, and start the backend server:

chmod +x run.sh  # (Only once)
./run.sh your_groq_api_key_here

This will:

Export the GROQ_API_KEY to your environment

Install Python dependencies

Start the backend server on http://127.0.0.1:5000

2. Manual Setup

In the env file Update

GROQ_API_KEY=your_groq_api_key_here

pip install -r requirements.txt
cd backend
python app.py

The server will start on http://127.0.0.1:5000 Follow the on-screen instructions for signup, login, and interacting with the system.

👤 Admin Access Instructions

By default, new users are registered as regular users. To make a user an admin, follow these steps:

After the user signs up, open the file: backend/user.json

Locate the user entry and modify their role like this: { "username": "your_username", "password": "hashed_password", "role_type": "admin" }

⚠️ Important: Modify role_type carefully to maintain system security. Admins may access sensitive features or data.