This is a Retrieval-Augmented Generation (RAG) application that allows users to upload documents and query them using natural language. The application uses Azure OpenAI for embeddings and language model capabilities, with a vector database for efficient document retrieval.

• FastAPI for the REST API
• LangChain and LlamaIndex for RAG capabilities
• Azure OpenAI for embeddings and LLM
• Weaviate as the vector database
• Unstructured.io for document parsing
• Temporal.io for workflow orchestration
• MinIO for object storage
• PostgreSQL for Temporal.io persistence

• Next.js 15.1 with TypeScript
• TailwindCSS for styling
• Dark/Light mode support
• File upload and chat interface

• Docker and Docker Compose
• Python 3.12+
• Node.js 18+
• Azure OpenAI API access

1. First, start the required infrastructure using Docker Compose:

cd build
docker compose up -d

This will start:

• Weaviate (Vector Database) - Port 8081
• Unstructured API (Document Processing) - Port 8800
• PostgreSQL (Database) - Port 5432
• Temporal (Workflow Engine) - Port 7233
• Temporal UI - Port 8233
• MinIO (Object Storage) - Ports 9000, 9001

2. Set up the backend:

cd backend

Create and activate a virtual environment

python -m venv .venv
source .venv/bin/activate # On Windows: .venv\Scripts\activate

Install dependencies

uv sync

Copy environment file and configure it

cp .env.dist .env

Edit .env with your Azure OpenAI and other credentials

Start the backend server

python -m uvicorn src.main:app --reload

3. Set up the frontend:

cd frontend

Install dependencies

npm install

Copy environment file and configure it

cp .env.dist .env

Set NEXT_PUBLIC_BACKEND_URL to your backend URL (default: http://localhost:8000)

Start the development server

npm run dev

4. Access the application:

• Frontend: http://localhost:3000
• Backend API: http://localhost:8000
• Temporal UI: http://localhost:8233
• MinIO Console: http://localhost:9001

• Document Upload: Upload and process various document types
• Two Query Modes:
  • Simple: Direct RAG with Azure OpenAI
  • Agentic: Advanced RAG with multi-step reasoning
• Source Attribution: See which documents were used to answer queries
• Dark/Light Theme Support
• Real-time Chat Interface

The application requires several environment variables to be set. Check the following files for required variables:

• Backend: backend/.env.dist
• Frontend: frontend/.env.dist

The project uses modern development tools and practices:

• Ruff for Python linting and formatting
• TypeScript for type-safe JavaScript
• TailwindCSS for styling
• Next.js App Router for frontend routing

This project is private and confidential. All rights reserved.