Marouf Chatbot - Intelligent Q&A Assistant 🤖

📌 Table of Contents

1. Overview
2. Competition Highlights
3. Key Features
4. Technical Stack
5. Installation
6. Usage
7. Architecture
8. Project Structure
9. Customization
10. Contact

🔍 Overview

Marouf Chatbot is an intelligent question-answering assistant leveraging RAG (Retrieval-Augmented Generation) architecture. It integrates FAISS for vector search, LLMs for response generation, and Redis caching to provide fast, accurate, and context-aware answers. Built with FastAPI and Streamlit, this chatbot ensures smooth interaction and seamless deployment via Docker.

🏆 Competition Highlights

• Innovative RAG Architecture combining LLMs with vector search.
• Optimized for low-latency (300ms avg response time).
• Modular design allowing easy model/dataset swaps.
• Production-ready deployment with Docker.

🌟 Key Features

🛠️ Technical Stack

Core Components

pie
title Technology Distribution
"NLP Processing" : 35
"Database" : 25
"API Server" : 20
"Frontend" : 15
"DevOps" : 5

🚀 Installation

# Clone repository
git clone https://github.com/Mkaljermy/marouf_chatbot.git
cd marouf_chatbot

# Setup environment
cp scripts/.env.example scripts/.env
nano scripts/.env  # Add your API keys

# Build and run
docker-compose up --build -d

💻 Usage

import requests

response = requests.post(
    "http://localhost:8000/chat",
    json={"query": "What's the capital of France?"}
)
print(response.json())

🏗️ Architecture

graph LR
    A[User] --> B[Streamlit]
    B --> C[FastAPI]
    C --> D{Redis?}
    D -->|Cache Hit| E[Return Response]
    D -->|Cache Miss| F[FAISS Search]
    F --> G[PostgreSQL]
    G --> H[LLM Processing]
    H --> C

📂 Project Structure

marouf_chatbot/
├── scripts/
│   ├── api/
│   │   ├── api.py
│   │   ├── embeddings.npy
│   │   └── faiss_index.index
│   ├── cache/
│   │   └── caching.py
│   ├── chatbot/
│   │   ├── chatbot.py
│   │   ├── embeddings.npy
│   │   └── faiss_index.index
│   └── frontend/
│       └── index.py
├── data/
│   └── trivia_dataset.csv
├── docker-compose.yml
├── Dockerfile.api
├── Dockerfile.frontend
└── requirements.txt

🛠️ Customization

• Modify Model: Replace Deepseek-Llama with another LLM by updating chatbot.py.
• Adjust Dataset: Replace trivia_dataset.csv in the data/ folder.
• Change UI: Edit frontend/index.py to update the Streamlit interface.

📞 Contact

For inquiries and contributions, contact Mohammad Aljermy:

• GitHub: @Mkaljermy
• LinkedIn: Mohammad Aljermy
• Email: kaljermy@gmail.com

🎉 Enjoy using Marouf Chatbot! 🚀

Check out the full project on GitHub: Marouf Chatbot Repository