Finance & ERP

Profile: financial · Model: DEFAULT (gpt-5.6-terra) · Chunking: table_aware · Confidence: ≥ 0.68

Overview

Parse invoices, purchase orders, financial statements, and SAP IDocs to extract companies, transactions, line items, amounts, and account codes for ERP integration, accounts payable automation, and financial analysis.

Extracted entities

Entity type	Examples
Company	Vendors, buyers, subsidiaries
Transaction	With reference number, type, and status
GLAccount	General ledger account codes
Amount	Monetary values with currency
InvoiceLineItem	Description, quantity, unit price, total
Date	Invoice date, due date, posting date

Extracted relationships

ISSUED_BY · RECEIVED_BY · POSTS_TO · REFERENCES · PAID_BY

Table-aware chunking

Financial documents are table-heavy — invoices with line items, statements with account balances. The financial profile uses table-aware chunking combined with table extraction to:

• Convert embedded tables to markdown
• Inject them as dedicated LLM context windows
• Auto-split large tables with header preservation
• Associate line items with their amounts, quantities, and descriptions

SAP IDoc support

DocIntel parses IDoc segment structure and maps INVOIC02 / ORDERS05 fields to the financial ontology automatically. No custom configuration needed.

Example

curl -X POST http://localhost:8200/process \
  -H "Authorization: Bearer $API_KEY" \
  -F "file=@invoice.pdf" \
  -F "profile=financial"

{
  "entities": [
    {"type": "Company", "name": "Acme Corp", "confidence": 0.74},
    {"type": "InvoiceLineItem", "name": "Widget A × 100", "confidence": 0.71},
    {"type": "InvoiceLineItem", "name": "Gadget B × 50", "confidence": 0.68},
    {"type": "Amount", "name": "$1,500.00", "confidence": 0.72},
    {"type": "Date", "name": "2026-04-01", "confidence": 0.69}
  ],
  "relationships": [
    {"source": "Acme Corp", "target": "$1,500.00", "type": "ISSUED_BY"}
  ]
}

Typical workflow

Invoices / POs / Statements (PDF, XLSX)
    ↓
DocIntel (financial profile) → companies, amounts, line items, dates
    ↓
JSON API / Webhook → structured financial data
    ↓
ERP system → auto-posting, three-way matching, AP automation