LLM & LLMOps

Chatbot LLMOps with RAG

A production-ready chatbot system implementing Retrieval-Augmented Generation (RAG) with Django, Celery, and OpenAI. The system uses Qdrant for vector storage and MariaDB for conversation management.

Architecture

• Backend: Django + Django REST Framework
• Vector Store: Qdrant
• Database: MariaDB
• Message Queue: Redis + Celery
• LLM Integration: OpenAI API
• Frontend: Streamlit UI

Features

• Conversational memory with persistent storage
• Asynchronous message processing using Celery
• Vector similarity search for relevant context retrieval
• Document management system
• Real-time chat interface
• Admin dashboard for system management

Prerequisites

• Python 3.10+
• Docker and Docker Compose
• OpenAI API key
• MariaDB
• Redis

Installation

1. Clone the repository

2. Set up environment variables:

env
MYSQL_ROOT_PASSWORD='root_password'
MYSQL_HOST='mariadb-tiny'
MYSQL_PORT=3306
OPENAI_API_KEY='your-openai-api-key'

3. Create network:

docker network create internal-network

4. Install dependencies:

For Mac:

bash
brew install mysql-client pkg-config
export PKG_CONFIG_PATH="$(brew --prefix)/opt/mysql-client/lib/pkgconfig"