Deterministic Invoice & Receipt Text-to-JSON Normalizer

A resilient, ultra-fast financial document parser that transforms unformatted OCR text dumps, email receipts, and digital invoice strings into standardized accounting JSON. Extracts merchant/vendor, invoice number, ISO 8601 date, currency, subtotal, tax/VAT, grand total, and itemized line items without calling expensive LLM APIs.

---

🚀 Why Use This Actor?

Using LLMs (like GPT-4o or Claude 3.5) for high-volume receipt and invoice intake costs $0.02 to $0.05 per document and suffers from token latency and non-deterministic formatting errors. This Actor provides a deterministic alternative:

• 100x Lower Costs: Billable at only $0.001 per invoice, cutting automated bookkeeping expenses by 95%+.
• Zero Hallucinations: Uses strict regex, financial token parsing, and currency heuristics. Numbers add up exactly as stated on the receipt.
• Automated ISO 8601 Date Parsing: Handles standard international date formats (YYYY-MM-DD, DD/MM/YYYY, MM/DD/YYYY, and month names).
• Line Item Extraction: Automatically extracts product descriptions, quantities, unit prices, and line totals.
• ERP & Accounting Integration: Feeds cleanly into QuickBooks, Xero, NetSuite, Google Sheets, or Airtable via Make/n8n.

---

📥 Input Schema

Field	Type	Default	Description
invoices	Array	Sample Receipt	Array of objects containing invoice_id and raw raw_text strings.
default_currency	String	"USD"	Fallback ISO currency code if no symbol is detected.

Sample Input JSON

{
  "invoices": [
    {
      "invoice_id": "INV-2026-01",
      "raw_text": "GitHub Inc.\nInvoice #GH-88192\nDate: 2026-10-01\nCopilot Enterprise  5  $39.00  $195.00\nSubtotal: $195.00\nTax: $0.00\nTotal: $195.00"
    }
  ],
  "default_currency": "USD"
}

---

📤 Output Dataset Format

{
  "invoice_id": "INV-2026-01",
  "vendor_name": "GitHub Inc.",
  "invoice_number": "GH-88192",
  "invoice_date_iso": "2026-10-01T00:00:00+00:00",
  "currency": "USD",
  "subtotal": 195.0,
  "tax_amount": 0.0,
  "total_amount": 195.0,
  "line_items": [
    {
      "description": "Copilot Enterprise",
      "quantity": 5,
      "unit_price": 39.0,
      "line_total": 195.0
    }
  ],
  "confidence_score": 100.0,
  "status": "success"
}

---

💰 Pay-Per-Event (PPE) Pricing

• Event Name: item-processed
• Cost: Only $0.001 per invoice ($1.00 per 1,000 invoices processed).
• Outcompetes manual data entry and expensive OCR SaaS APIs with instantaneous throughput and zero setup overhead.

---

🔌 API & Python Client Example

from apify_client import ApifyClient

client = ApifyClient("YOUR_APIFY_TOKEN")

run = client.actor("YOUR_USERNAME/invoice-receipt-json-normalizer").call(run_input={
    "invoices": [
        {"invoice_id": "receipt_1", "raw_text": "Office Depot\nTotal: $45.20\nDate: 2026-08-14"}
    ]
})

for item in client.dataset(run["defaultDatasetId"]).iterate_items():
    print(f"Vendor: {item['vendor_name']} | Total: {item['currency']} {item['total_amount']}")