Document Q&A Demo

A document question answering system with FastAPI backend and React frontend that allows you to upload documents (PDF, DOCX, TXT) and ask natural language questions about them.

Features

• Document upload and processing (PDF, DOCX, TXT)
• Document chunking and embedding generation
• Hybrid search (vector + keyword) for accurate information retrieval
• OpenAI embeddings with e5 model fallback
• React UI for document upload and question asking

Project Structure

doc-qa-demo/
│
├─ api/               # FastAPI code
│   ├─ ingest.py      # parses & chunks docs
│   ├─ embed.py       # wraps OpenAI + fallback e5
│   ├─ retrieve.py    # hybrid search
│   └─ main.py        # /ask endpoint
└─ web/               # React UI

Setup Instructions

Backend (FastAPI)

1. Create and activate a Python virtual environment:

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

2. Install required packages from requirements.txt to avoid dependency conflicts:

   cd api
   pip install -r requirements.txt

   Note: If you're using TensorFlow, there may be compatibility issues with protobuf versions. The requirements.txt file specifies protobuf<5.0.0 to ensure compatibility with TensorFlow.

3. Set your OpenAI API key as an environment variable:

   export OPENAI_API_KEY=your-api-key-here  # On Windows: set OPENAI_API_KEY=your-api-key-here

4. Run the FastAPI server:

   uvicorn main:app --reload

   The API will run on http://localhost:8000

Frontend (React)

1. Install Node.js dependencies:

   cd web
   npm install

2. Run the React development server:

   npm start

   The UI will run on http://localhost:3000

Usage

1. Open http://localhost:3000 in your browser
2. Upload one or more documents using the document uploader
3. Ask questions about the uploaded documents
4. View the results showing the most relevant document chunks

API Endpoints

• GET /: Root endpoint to check if API is running
• POST /documents/upload: Upload a document for processing
• POST /ask: Ask a question about the documents
• GET /documents: List all uploaded documents

Requirements

• Python 3.8+
• Node.js 14+
• OpenAI API key (optional, e5 model will be used as fallback)

Troubleshooting

Dependency Conflicts

If you encounter dependency conflicts with protobuf and TensorFlow, you can fix them by installing specific versions:

pip uninstall protobuf -y
pip install protobuf>=3.20.3,<5.0.0

For other dependency issues, try installing the requirements.txt file with the --no-deps flag and then manually installing the conflicting packages:

pip install -r requirements.txt --no-deps
pip install package-name==specific-version