Overview

• Cancelled cheque
• Aadhaar card
• PAN card
• Invoice
• Driving license
• Vehicle RC
• Passport
• Voter ID

​

Use cases

• KYC verification: Instantly extract and verify customer details from official documents for onboarding or compliance.
• Loan applications: Validate income and identity documents in real-time during loan processing.
• eCommerce and fintech platforms: Speed up seller or partner registration by auto filling details from uploaded documents.
• Gig economy platforms: Automate the document submission and verification flow for contractors or service partners.
• Travel and hospitality: Document Verification of Passport, VISA OCR for travel bookings.

​

Key benefits

• Single integration: Integrate once to handle different document types without managing multiple APIs. Smart OCR detects blurry or glary images, flags partial or obscured IDs, and prompts users to retake poor-quality images with clear feedback. This reduces rejections and lowers drop-offs.
• Faster time to market: Launch your verification process faster with less development and testing effort.
• Cost efficiency: Reduce maintenance costs with a single, streamlined API instead of managing multiple integrations.
• Reliable accuracy: Smart OCR uses industry-grade optical character recognition for high accuracy in text extraction and validation. It achieves over 97% accuracy across supported documents, extracts structured data, and works for all document types through one API.
• Advanced fraud and quality checks: Smart OCR automatically checks document quality and authenticity to prevent fraud and ensure reliable extraction. It detects blur, glare, partial images, black-and-white scans, and missing facial or QR information. It also flags screenshots and photos of screens to block tampered or fake documents.
  Upcoming Enhancement:

  • Document Auto Classification: Automatically identifies and categorises the document type eliminating the need to manually specify document types.

​

Verification process

• Merchant dashboard: Manual verification with upload interface for quick document processing.
• REST API: Programmatic integration for automated document verification workflows.

• Dashboard verification
• API verification

Use the merchant dashboard for quick manual Smart OCR verification:

​

Smart OCR verification

1. Log in to the Merchant Dashboard.
2. Navigate to KYC Studio > Smart OCR.
3. Select the document type from the dropdown (PAN card, Aadhaar card, Voter ID, Vehicle RC, Driving Licence, Passport).
4. Upload the document image (maximum 5 MB, supported formats: .png, .jpeg, .jpg).
5. Click Verify to start the OCR process.

Upload guidelines:

• Maximum file size: 5 MB
• Supported formats: .png, .jpeg, .jpg
• Upload the side containing critical information (photo and ID number)
• Ensure clear, well-lit images without glare or shadows

The dashboard displays extracted data, validation results, and document quality analysis upon completion.

Integrate Smart OCR verification programmatically using the REST API:

1. Trigger the Smart OCR request using the Smart OCR API.
2. Pass the document image and document type in the request payload for processing.
3. Receive OCR results with extracted data, validation status, and quality analysis.

Endpoint: POST /verification/smart-ocr

Request format

{
  "verification_id": "OCR00123",
  "document_type": "PAN_CARD",
  "image": "base64_encoded_image_string"
}

Response format

{
  "verification_id": "OCR00123",
  "reference_id": 12345,
  "document_type": "PAN_CARD",
  "extracted_data": {
    "pan_number": "ABCDE1234F",
    "name": "John Doe",
    "father_name": "Robert Doe",
    "date_of_birth": "01/01/1990"
  },
  "quality_checks": {
    "image_quality": "HIGH",
    "blur_detection": "PASS",
    "glare_detection": "PASS",
    "document_completeness": "COMPLETE"
  },
  "fraud_checks": {
    "screenshot_detection": "PASS",
    "tampering_detection": "PASS",
    "authenticity_score": 0.95
  },
  "verification_status": "SUCCESS"
}

Error responses

{
  "type": "invalid_request_error",
  "code": "ocr_failed",
  "message": "Document processing failed",
  "details": [
    {
      "field": "image",
      "issue": "Image quality too low for processing"
    }
  ]
}