kyc-extractor-ai 0.1.0

AI-powered KYC Document Extraction using Google Gemini

Company Name Cleaning (CC) - POC

This project is a Proof of Concept (POC) for automatically extracting company details from KYC documents (GST, PAN, FSSAI, etc.) using Google Gemini Flash (Vision LLM).

Features

• Multi-Format Support: Handles PDF (single & multi-page) and Images (JPG, PNG).
• Intelligent Extraction: Uses Gemini 2.5 Flash to extract:
  • Company Name (Legal vs Trade Name)
  • Address (Bifurcated into Line 1, Locality, City, State, Pincode)
  • Identification Numbers (GSTIN, PAN, etc.)
  • Dates & Approver Names
• Confidence Scoring: Returns a confidence score (0-1) and a reason for the score.
• API: Fast, async-ready API built with FastAPI.

Prerequisites

• Python 3.10+
• poppler-utils (for PDF processing)
  • Ubuntu: sudo apt-get install poppler-utils
  • Mac: brew install poppler
• Google Gemini API Key

Setup

1. Clone the repository

   git clone <repo-url>
   cd cc-poc
   

2. Install Dependencies

   pip install -r requirements.txt
   

3. Environment Variables Create a .env file in the root directory:

   GOOGLE_API_KEY=your_actual_api_key_here
   

Usage

Running the API Server

uvicorn kyc_extractor.main:app --host 0.0.0.0 --port 8000 --reload

Testing the API

You can use the provided test script:

python scripts/test_api.py /path/to/document.pdf

Or use curl:

curl -X POST "http://localhost:8000/extract" \
     -H "accept: application/json" \
     -H "Content-Type: multipart/form-data" \
     -F "file=@/path/to/document.pdf"

Project Structure

• kyc_extractor/: Core application code.
  • main.py: API endpoints.
  • core/: Config, Gemini client, and Prompts.
  • services/: Image processing logic.
  • schemas.py: Pydantic data models.
• docs/: Documentation (PRD, Tech Doc, Plans).
• scripts/: Utility scripts for testing.

Docker Support

Build and run with Docker Compose:

docker-compose up --build