DocuQuery - AI-Powered Document Q&A

An AI-powered PDF document analysis application with Retrieval-Augmented Generation (RAG). Chat with your PDFs using OpenAI-compatible APIs with multimodal content extraction and semantic search.

Features

• PDF processing with text, table, and image extraction
• Semantic search using FAISS vector store with sentence transformers
• RAG-based question answering with citations
• Modern React frontend with dark mode
• Flexible API configuration (OpenAI, NVIDIA, or any compatible API)
• Runtime API settings through UI without restarting
• Conversation management with multi-turn support

Tech Stack

Backend:

• FastAPI (Python web framework)
• SQLAlchemy (database ORM)
• FAISS (vector similarity search)
• Sentence Transformers (embeddings)
• PyMuPDF & pdfplumber (PDF processing)
• OpenAI client (compatible with any OpenAI-like API)

Frontend:

• React 18
• Tailwind CSS
• Lucide React (icons)
• Axios (HTTP client)

Quick Start

Prerequisites

• Python 3.8+
• Node.js 14+
• Git

Installation

1. Clone the repository:

git clone https://github.com/devJinesh/DocuQuery.git
cd DocuQuery

2. Backend setup:

cd backend
python -m venv venv

# Windows:
venv\Scripts\activate
# Mac/Linux:
source venv/bin/activate

pip install -r requirements.txt

# Create .env file
copy .env.example .env  # Windows
cp .env.example .env    # Mac/Linux

3. Configure your .env file:

APP_NAME=DocuQuery
DATABASE_URL=sqlite+aiosqlite:///./data/docuquery.db

# LLM Configuration
OPENAI_API_KEY=your_api_key_here
OPENAI_BASE_URL=https://integrate.api.nvidia.com/v1
MODEL_NAME=nvidia/llama-3.1-nemotron-70b-instruct

# Optional
ALLOWED_ORIGINS=http://localhost:3000
MAX_FILE_SIZE_MB=50

4. Start the backend:

python -m uvicorn main:app --reload --host 0.0.0.0 --port 8000

5. Frontend setup (new terminal):

cd frontend
npm install
npm start

The application will open at http://localhost:3000

Using the Application

1. Upload a PDF: Click the "Upload" tab and select a PDF file
2. Chat: Select a document from "Documents" tab, switch to "Chat" tab, and ask questions
3. Configure API (optional): Click the settings icon to customize API endpoint, key, and model
4. Dark Mode: Toggle with the moon/sun icon in the header

Development

Quick start scripts are available:

Setup (first time only):

setup.bat    # Windows
./setup.sh   # Mac/Linux

Start application:

start.bat    # Windows
./start.sh   # Mac/Linux

License

MIT License - feel free to use for personal or commercial projects.