FastAPI with BERT Document Analyzer

An intelligent Document Analysis API built using FastAPI, Hugging Face Transformers, and PyTorch. The system leverages a custom fine-tuned BERT model to classify text extracted from multiple document formats. It supports raw text, PDFs, Microsoft Word documents, Excel spreadsheets, CSV files, and images using Optical Character Recognition (OCR).

---

✨ Features

• 🤖 Custom fine-tuned BERT text classification model
• 🚀 High-performance FastAPI REST API
• 📝 Analyze raw text input
• 📄 PDF document processing
• 📘 Microsoft Word (.docx) support
• 📊 Excel (.xls/.xlsx) and CSV analysis
• 🖼️ OCR-based text extraction from images
• 📈 Prediction confidence scores
• 📚 Interactive API documentation (Swagger & ReDoc)
• ⚡ Automatic model loading during application startup
• 🔄 Supports multiple document formats through a single API

---

📂 Supported File Formats

File Type	Extension	Supported
Text	.txt	✅
PDF	.pdf	✅
Word	.docx	✅
Excel	.xlsx	✅
Excel	.xls	✅
CSV	.csv	✅
Image	.png	✅
Image	.jpg	✅
Image	.jpeg	✅

---

📁 Project Structure

FastAPI-BERT-Document-Analyzer/
│
├── compliance/
│   └── rule_engine.py
│
├── intent/
│   └── scoring.py
│
├── nlp/
│   ├── fastapi_bert.py        # FastAPI application
│   └── bert_model.py          # BERT model training
│
├── my_custom_model/
│   └── final/                 # Saved trained model
│
├── requirements.txt
├── README.md
└── LICENSE

---

⚙️ Installation

1. Clone the Repository

git clone https://github.com/yourusername/FastAPI-BERT-Document-Analyzer.git

cd FastAPI-BERT-Document-Analyzer

---

2. Create a Virtual Environment

Windows

python -m venv venv

venv\Scripts\activate

Linux / macOS

python3 -m venv venv

source venv/bin/activate

---

3. Install Dependencies

pip install -r requirements.txt

---

🏋️ Train the BERT Model

Run the training script:

python nlp/bert_model.py

After successful training, the model will be saved inside:

my_custom_model/final/

---

🚀 Run the FastAPI Server

Start the API using Uvicorn:

uvicorn nlp.fastapi_bert:app --reload

The application will be available at:

http://127.0.0.1:8000

---

📚 API Documentation

Swagger UI

http://127.0.0.1:8000/docs

ReDoc

http://127.0.0.1:8000/redoc

---

📌 API Endpoints

1️⃣ Analyze Raw Text

POST

/analyze/text

Request

{
    "text": "This document looks suspicious."
}

Response

{
  "label": "Positive",
  "confidence": 98.74
}

---

2️⃣ Analyze Document

POST

/analyze/file

Upload any supported document.

Supported formats:

• TXT
• PDF
• DOCX
• XLS
• XLSX
• CSV
• PNG
• JPG
• JPEG

Example Response

{
    "filename": "report.pdf",
    "file_type": "pdf",
    "analysis": {
        "label": "Positive",
        "confidence": 98.74
    }
}

---

🧠 Processing Workflow

                User Input
                     │
      ┌──────────────┴──────────────┐
      │                             │
   Raw Text                    Uploaded File
      │                             │
      │                     Detect File Type
      │                             │
      │      ┌──────────────┬──────────────┐
      │      │              │              │
      │     PDF          DOCX/XLSX      Images
      │      │              │              │
      │      └──────Extract Text───────────┘
      │                    │
      └──────────────► Preprocessing
                           │
                           ▼
                 Fine-Tuned BERT Model
                           │
                           ▼
                 Classification Result
                           │
                           ▼
                JSON Response (Label + Confidence)

---

🛠️ Technology Stack

• Python
• FastAPI
• Hugging Face Transformers
• PyTorch
• Pandas
• NumPy
• Scikit-learn
• Pydantic
• PyPDF2
• python-docx
• Pillow
• pytesseract
• OpenPyXL
• Uvicorn

---

📈 Future Enhancements

• 🌍 Multi-language document classification
• 📂 Batch document processing
• 📄 Long document chunking
• ⚡ GPU acceleration
• 🐳 Docker support
• ☸️ Kubernetes deployment
• 🔐 JWT Authentication
• 📦 Model versioning
• 🔄 CI/CD Pipeline
• ☁️ AWS, Azure & Google Cloud deployment
• 📊 Model monitoring and analytics

---

🤝 Contributing

Contributions are welcome!

1. Fork this repository.

2. Create a new branch.

git checkout -b feature-name

3. Commit your changes.

git commit -m "Add new feature"

4. Push to GitHub.

git push origin feature-name

5. Open a Pull Request.

---

📄 License

This project is licensed under the MIT License.

---

👨‍💻 Author

Muhammad Hamid

---

⭐ If you found this project helpful, consider giving it a Star on GitHub!