Metadata Extraction API

Overview

This project provides a REST API to extract contract metadata such as:

• Award Criteria
• Solvency Criteria
• Special Execution Conditions

The extraction is based on text documents using OpenAI's GPT-4o-mini model.
The API processes raw text input and returns a structured JSON output with the extracted information.

---

✅ Features

• 📡 REST API for contract metadata extraction.
• 🤖 Powered by OpenAI GPT-4o-mini.
• 📄 Supports plain text as input.
• 🧠 Automatic extraction of 3 categories:
  • Award Criteria
  • Solvency Criteria
  • Special Execution Conditions
• 📜 Logging system.
• 🐳 Dockerized for easy deployment.

---

📂 Project Structure

.
├── app
│   ├── main.py               # Main Flask application
│   └── requirements.txt      # Python dependencies
├── .env                      # Environment variables (API key)
├── .gitignore                # Git ignore file
├── Dockerfile                # Docker build file
├── docker-compose.yml        # Docker Compose setup
└── README.md                 # This documentation

---

⚙️ Setup & Installation

1️⃣ Clone the Repository

git clone https://github.com/your-repo/metadata-extraction-api.git
cd metadata-extraction-api

2️⃣ Set OpenAI API Key

Create a .env file in the root directory and add your OpenAI API key:

OPENAI_API_KEY=sk-your-api-key

You can obtain an API key by registering at OpenAI's API platform.

---

🐳 Run with Docker

1️⃣ Build the Docker Image

docker build -t metadata-api .

2️⃣ Run the API in a Container

docker run -p 5000:5000 --env-file .env metadata-api

3️⃣ Run with Docker Compose (recommended)

docker-compose up --build

This will expose the API at http://localhost:5000/extract_metadata.

---

🚀 API Usage

Endpoint

POST /extract_metadata

Request Body (JSON)

Field	Type	Description
text	string	Plain text content to analyze

Example Request (cURL)

curl -X POST "http://localhost:5000/extract_metadata" -H "Content-Type: application/json" -d '{
  "text": "This contract is awarded based on financial and technical criteria..."
}'

Example Response (JSON)

{
  "criterios_adjudicacion": "Award criteria extracted...",
  "criterios_solvencia": "Solvency criteria extracted...",
  "condiciones_especiales": "Special conditions extracted..."
}

---

⚠️ Input Format Requirements

• The text field must contain plain text.
• The API automatically processes large text inputs by splitting them into smaller sections before analysis.

---

📜 Logging

• Logs are printed directly to the console when running the container.
• You can extend logging to file inside app/main.py if needed.

---

🛑 Stopping and Removing Containers

To stop and remove the containers:

docker-compose down

To prune unused images and containers:

docker system prune -a

---

✅ Example .env file

OPENAI_API_KEY=sk-your-openai-api-key

---

🚀 Run without Docker (Optional)

If you want to run it locally without Docker:

1️⃣ Create a Virtual Environment

python3 -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

2️⃣ Install Dependencies

pip install -r app/requirements.txt

3️⃣ Run the Flask App

python app/main.py

API will be available at http://localhost:5000/extract_metadata.

---

This version is optimized for direct text-based metadata extraction and supports large text inputs efficiently. 🚀

Acknowledgements

This work has received funding from the NextProcurement European Action (grant agreement INEA/CEF/ICT/A2020/2373713-Action 2020-ES-IA-0255).