DockConflictDetector

This project is a FastAPI application that detects conflicting statements in PDF documents. It uses Azure OpenAI service to analyze potential conflicts within PDF content.

Requirements

• Python 3.8+
• FastAPI
• PyPDF2
• Azure OpenAI API access

Installation

1. Clone the project:

git clone <repository-url>
cd DockConflictDetector

2. Create and activate virtual environment:

python -m venv venv
source venv/bin/activate  # Unix/macOS
# or
.\venv\Scripts\activate  # Windows

3. Install required packages:

pip install -r requirements.txt

4. Create .env file and add required variables:

OPENAI_API_KEY=your_api_key
OPENAI_API_VERSION=your_api_version
OPENAI_API_BASE=your_api_base
OPENAI_DEPLOYMENT_NAME=your_deployment_name

Running the Application

To start the application:

uvicorn app.main:app --reload --port 8000

To access the API documentation, open the following URL in your browser:

http://localhost:8000/docs

API Usage

• Use the /upload endpoint to upload PDF files
• Only PDF format files are accepted
• The content of the uploaded PDF is analyzed to detect conflicting statements
• Results are returned in JSON format