AI Data Extraction Tool

🚀 Upload documents → Extract Names, Emails, and Organizations → Download structured Excel results instantly.
Built with FastAPI, Pandas, and optional GPT-enhanced extraction.
Deployed live on Render.

---

✨ Features

• ✅ Upload PDF, DOCX, and TXT documents
• ✅ Extract Names, Emails, and Organizations
• ✅ Multi-file uploads supported (combines results into one Excel)
• ✅ Clean and organized Excel file download (.xlsx)
• ✅ Supports both local entity extraction and GPT-enhanced extraction
• ✅ Automatic fallback if the custom model is missing
• ✅ Deployed online via Render

---

📸 Screenshots

Upload Page

Extraction Results Page

---

🚀 Live Demo

🟢 Visit the Live App Here

---

⚙️ Technologies Used

• Python 3.11
• FastAPI
• Uvicorn
• Pandas
• spaCy
• OpenAI API (optional GPT-enhancement)
• openpyxl (for Excel export)

---

🛠 Local Development Setup

Clone the repository:

git clone https://github.com/PMTheTechGuy/document-entity-extractor.git
cd document-entity-extractor

Install dependencies:

pip install -r requirements.txt

Set up your environment variables:

Create a .env file based on .env.example.

cp .env.example .env

Start the server locally:

uvicorn api.main:app --reload

If you encounter an issue loading the application on HTTP://localhost:8000.

Quit the application using Ctrl + C and start the server on port 8001.

uvicorn api.main:app --reload --port 8001

---

🧠 OpenAI Key Setup (Optional for GPT Extraction)

This app supports two extraction modes:

• 🧠 GPT-enhanced extraction (more accurate, slower, uses OpenAI API)

• ⚡ Local spaCy model extraction (faster, free, no external API calls)

By default, the app will fall back to spaCy if no OpenAI key is provided and the USE_GPT_EXTRACTION is set to False.

Setting Up OpenAI GPT Extraction (Optional)

1. In your .env file, add your OpenAI API Key:

OPENAI_API_KEY=your-real-openai-api-key-here

2. Save the .env file.

3. Restart the FastAPI server:

uvicorn api.main:app --reload

• ✅ If a key is provided, the app will automatically use GPT for extractions.
• ✅ If no key is provided or an API error occurs, the app will fall back to using spaCy.

---

⚙️ Controlling GPT Extraction Mode

In your .env file, you can control whether the app uses GPT or local spaCy extraction:

USE_GPT_EXTRACTION=True

• ✅ True → Use GPT extraction (requires valid OpenAI API key)

• ✅ False → Force local spaCy extraction, even if API key is present

Restart the server after changing the .env settings.

uvicorn api.main:app --reload

The app will detect this automatically at runtime.

---

🌍 Deployment

This app is deployed on Render.

You can deploy your version in one click.

---

📦 Folder Structure

api/             # FastAPI backend
├── templates/   # HTML templates (upload form, results page)
├── static/      # Static files
utils/           # Helper modules (export, logging, etc.)
extractor/       # File reading and entity extraction
gpt_integration/ # GPT-enhanced extraction
output/          # Exported Excel files
logs/            # Application logs

---

📦 Features

• Multi-file Upload: Upload one or more .pdf, .docx, or .txt files for processing.
• Entity Extraction: Automatically identifies and extracts:
  • People (names)
  • Emails
  • Organizations
• Results Summary: Displays a summary of total files processed, and the number of names, emails, and organizations found.
• CSV & Excel Export: Download extracted data in .csv or .xlsx format.
• Auto Cleanup: Temporary files that are older than one hour will be automatically deleted.
• Error Handling: User interface for handling invalid uploads, unsupported file types, and extraction failures.

---

🚧 Coming Soon

• Daily upload limits per user or IP (via database tracking)
• Admin dashboard to review processed data
• File size limit configuration in .env

---

🙌 Acknowledgements

• FastAPI

• spaCy

• OpenAI

• Render

---

📫 Contact

Crafted with dedication by PM The Tech Guy.

Please don't hesitate to reach out or share your ideas!

---

📝 License

This project is licensed under the MIT License.