Document OCR

Extract data from any document in seconds

Payroll receipts, bank statements, proof of address and more. Our AI reads, interprets and structures information automatically.

Documents we process

Payroll receipts

Gross and net salary, deductions, pay period, employer and employee RFC.

Bank statements

Average balance, income, expenses, account number, issuing bank and period.

Proof of address

Full address, account holder name, service type and issue date.

ID documents

Name, CURP, date of birth, address and photograph from INE/IFE.

How does it work?

Step 1

Upload the document

Send the image or PDF of the document. We accept JPG, PNG and PDF.

Step 2

AI processes and extracts

Our models identify the document type and extract every field automatically.

Step 3

Receive structured data

Clean JSON with every field identified, ready to use in your system.

ocr-extract.sh
curl -X POST https://api.singula.mx/v1/ocr/extract \
  -H "Authorization: Bearer sk_live_..." \
  -F "document=@recibo_nomina.pdf" \
  -F "customer_id=cus_lzk9m4ab3f"

→ { "status": "processing", "webhook": true }

// Webhook:
{
  "document_type": "payroll_receipt",
  "extracted": {
    "gross_salary": 45000,
    "net_salary": 38500,
    "employer_rfc": "TEC900101AB1",
    "employee_rfc": "GARM900315AB1",
    "period": "2025-03",
    "deductions": 6500
  },
  "confidence": 0.97,
  "fields_extracted": 6
}

Multiple formats

JPG, PNG and PDF. Scanned documents or phone photos.

Automatic classification

No need to specify the type. AI identifies what document it is.

Cross-validation

Extracted data is cross-checked with CURP, RFC and others for consistency.

Alteration detection

Identifies digitally edited, cropped, or manipulated documents.

Async with webhook

Processing takes seconds. We notify you when it's ready.

Ideal for onboarding

Validate income and address as part of the customer onboarding flow.

Digitize documents with intelligence

No contracts. Free sandbox. Production ready.

Create free account