📄 Data Extraction API

---

🧠 Description

This project is an AI-powered Document Analyzer API built using FastAPI. It processes documents (PDF, DOCX, and images) by extracting text, generating summaries, identifying key entities, and performing sentiment analysis.

The system accepts files in Base64 format, decodes them, extracts content using OCR or parsing techniques, and applies NLP-based analysis to return structured insights.

---

⚙️ Tech Stack

• Language/Framework: Python, FastAPI
• Key Libraries: Uvicorn, Pydantic, Base64, pytesseract, PIL
• AI/NLP: Basic NLP techniques for summarization and sentiment analysis

---

🏗️ Architecture Overview

The system follows a simple API-based architecture:

Client (Postman / Swagger) ↓ FastAPI Backend (Render) ↓ Document Processing Layer ↓ Text Extraction (PDF/DOCX/OCR) ↓ NLP Analysis (Summary, Entities, Sentiment) ↓ JSON Response

🛠️ Setup Instructions

1. Clone the repository

git clone https://github.com/prav-tech/document-analyzer-api.git
cd document-analyzer-api

2. Install dependencies

pip install -r requirements.txt

3. Set environment variables Create a .env file:

API_KEY=your_secret_key

4. Run the application

uvicorn src.main:app --reload

5. Open in browser:

http://127.0.0.1:8000/docs

---

🔍 Approach

1. Input Handling

   • Accept document as Base64 string
   • Identify file type (PDF, DOCX, Image)

2. File Processing

   • Decode Base64 to file
   • Save temporarily

3. Text Extraction

   • PDFs/DOCX → direct parsing
   • Images → OCR using Tesseract

4. Analysis

   • Generate summary
   • Extract entities (names, dates, organizations, etc.)
   • Perform sentiment analysis

5. Output

   • Return structured JSON response

---

🚀 Live API

• https://quickai-qy5b.onrender.com
• Swagger Docs: https://quickai-qy5b.onrender.com/docs

---

📦 API Endpoint

POST /api/document-analyze

Headers:

x-api-key: your_secret_key
Content-Type: application/json

Request Body:

{
  "fileName": "sample.pdf",
  "fileType": "pdf",
  "fileBase64": "BASE64_STRING"
}

📤 Sample Response

{
  "status": "success",
  "fileName": "sample.pdf",
  "summary": "Short summary...",
  "entities": {
    "names": [],
    "dates": [],
    "organizations": [],
    "amounts": [],
    "locations": []
  },
  "sentiment": "Neutral"
}

📌 Status

✅ API working ✅ Deployed on Render ✅ Tested using Postman & Swagger

🤖 AI Tools Used

• OCR: Tesseract (for image text extraction)
• NLP: Basic rule-based and text processing techniques
• AI Assistance: ChatGPT (for debugging, guidance, and optimization)

⚠️ Known Limitations

• Entity extraction may not always detect all entities accurately
• Sentiment analysis is basic and may not reflect deep context
• OCR accuracy depends on image quality
• Large files may take longer to process
• No frontend UI (API-based interaction only)

👩‍💻 Author

Pravallika B