Home / Patterns / Document Completeness Check Failures

Document Completeness Check Failures

Goal Document Verification Frequency Common Category Mortgage Documents Published View source on GitHub ↗

Issue: OCR System Fails to Detect Missing Pages, Sections, or Required Information

Frequency: Common

Symptoms

  • Missing pages in multi-page documents not flagged
  • Blank pages not identified
  • Required fields empty but document “processed”
  • Truncated documents accepted
  • Attachments/exhibits missing
  • Version mismatches undetected

Root Cause Mortgage document packages contain hundreds of pages across dozens of documents. OCR must verify completeness—all pages present, required fields populated, attachments included. Missing or incomplete documents create compliance issues, closing delays, and potential fraud opportunities.

Example

Scenario 1: Missing appraisal pages

Appraisal submitted: Pages 1, 2, 3, 7, 8
Missing: Pages 4-6 (comparable sales photos)

OCR: Extracted data from available pages
Result: Property value extracted, but comparables incomplete

← Critical valuation data missing
← No page count validation
← Investor would reject

---

Scenario 2: Blank required fields

1003 Application:
- Employer name: "ABC Corp"
- Employer address: [blank]
- Employment start date: [blank]
- Phone number: [blank]

OCR: Extracted employer name ✓
Reality: Missing required employment verification data

---

Scenario 3: Exhibit reference without attachment

Deed of Trust states: "See Exhibit A for legal description"
Exhibit A: Not included in document package

← Legal description missing
← Reference detected but attachment not verified
← Title defect

---

Completeness check failures:
  
  Documents with completeness issues: 15%
  
  Issue types:
    Missing pages: 30%
    Blank required fields: 35%
    Missing attachments: 20%
    Truncated documents: 10%
    Wrong version: 5%
  
  Impact:
    Re-submission required: 12%
    Closing delays: 8%
    Compliance findings: 5%

Key Statistics From Document Processing Research (2026):

  • Incomplete document packages: 15-20%
  • Missing pages: 5-8% of documents
  • Blank required fields: 10-15%
  • Re-submission rate: 10-15%

Contributing Factors

  • No page count validation
  • Required field logic missing
  • Attachment references not traced
  • Version control absent
  • Scanning/upload errors undetected

Mitigation Strategies

Prevention

  1. Page count validation: Expected vs. actual pages
  2. Required field checks: Document-specific field requirements
  3. Attachment tracing: Verify all referenced exhibits
  4. Version validation: Correct form version for loan type
  5. Quality checks: Blank/corrupt page detection

Implementation

class DocumentCompletenessChecker:
    """Verify document completeness"""
    
    EXPECTED_PAGES = {
        "uniform_residential_appraisal": 8,
        "closing_disclosure": 5,
        "loan_estimate": 3,
        "1003_application": 5
    }
    
    REQUIRED_FIELDS = {
        "1003_application": [
            "borrower_name", "ssn", "employer_name",
            "employer_address", "income", "assets"
        ]
    }
    
    def check_page_count(self, doc_type: str, pages: int) -> dict:
        expected = self.EXPECTED_PAGES.get(doc_type)
        if expected and pages < expected:
            return {
                "complete": False,
                "expected": expected,
                "actual": pages,
                "missing": expected - pages
            }
        return {"complete": True}
    
    def check_required_fields(self, doc_type: str, fields: dict) -> dict:
        required = self.REQUIRED_FIELDS.get(doc_type, [])
        missing = [f for f in required if not fields.get(f)]
        
        return {
            "complete": len(missing) == 0,
            "missing_fields": missing
        }

References

Related Patterns