Release Notes

In this release, we’ve expanded support for additional Classify and Capture documents, introduced enhancements to the Docs-to-Digital feature (including severity scoring for mismatches, reason codes, related transaction linking, and a new Docs-to-Digital Excel sheet), added the new book.income.updated webhook event, and announced the deprecation of the Middesk integration.

What's new?

Webhook

New book.income.updated webhook event

A new event,`book.income.updated, has been added to both organization-level and account-level webhooks. This event is triggered whenever the income calculation for a specific Book is edited or updated. It ensures that systems and stakeholders are promptly informed of any changes to income data, supporting accurate downstream processing and reporting.

To learn more, see the Webhook page.

{
    "severity": "HIGH",
    "book_uuid": "a2801064-e875-4d1f-b43d-dbf5dfc907be",
    "event_name": "book.income.updated",
    "notification_type": "STATUS",
    "notification_reason": "Analyze Book Income Updated for the book"
}

Classify and Capture

New Classify and Capture supported documents

The Classify and Capture supports are enabled for the following newly onboarded documents:

Other

• Purchase Advice
• Purchase Advice and Wire Details (Final)

To learn more, see the All Supported Documents page.

Classify

New Classify-specific documents support

Classify support is now extended to the various documents. Click the category name below to expand/collapse the list of newly supported documents:

To learn more, see the All Supported Documents page.

Enhancements

Docs-to-digital matching - severity scoring

We’re excited to announce an enhancement to the Docs-to-Digital reconciliation workflow: severity scoring for mismatched transactions! This powerful new capability brings clarity and prioritization to the process of comparing Plaid transaction data with uploaded bank statements.

Underwriters often face the challenge of manually reviewing long lists of mismatches, many of which are low-risk. With this update, each mismatch is automatically assigned a severity level (High, Medium, or Low) and helps teams focus their attention on the most critical discrepancies, like NSFs, debt-related transactions, or potential fraud indicators.

To further support underwriting workflows, we’ve introduced a dedicated Docs-to-Digital Excel sheet in the SMB Excel export. This new sheet offers an at-a-glance summary of mismatch severity, key metrics comparing Plaid and bank statement data, and a detailed breakdown of each mismatched transaction, including reason codes, tags, and related transaction references. The result is a more efficient and consistent reconciliation process, enabling faster funding decisions with less manual effort.

Severity scoring for mismatched transactions

Docs-to-Digital now includes an intelligent scoring system that assigns a severity level to each mismatched transaction between Plaid data and uploaded bank statements. Mismatched transactions are automatically categorized as High, Medium, or Low severity based on business rules developed in collaboration with lending partners. High-severity mismatches occur when transactions involving NSFs, MCAs, overdrafts, refunds, crypto, or gambling appear in either the digital data source or the bank statement, but not both. Identifying these high-risk transaction mismatches helps underwriters focus on red flags that could impact funding decisions.

Example node in the enriched transaction endpoint :

"mismatched_transactions": [
  {
    "txn_pk": 123,
    "txn_date": "2024-01-01",
    "description": "TRANSFER TO CHECKING",
    "amount": -2000.01,
    "uploaded_doc_pk": 123456,
    "uploaded_doc_format": "PLAID",
    "severity": "High",
    "reason_code": "critical_category_mismatch",
    "related_txn_pk": 456,
    "tags": "fintech, fintech_loan",
    "counterparty": "Credibly"
  },
]

Book-level severity scoring

Every Book is now scored based on the most severe mismatch it contains. This new field, book_mismatch_severity in the Book summary API allows underwriters to triage Books at a glance. A Book with even one high-severity mismatch is flagged as High, making prioritization during high-volume periods more efficient and consistent. The book_mismatch_severity is exposed at 3 places: Book summary endpoint, Webhook response, and SMB exported excel sheet.

• Example of Book Summary endpoint where Book Mismatch Severity is found:

  JSON

  {
  "book_pk": 59582262,
  "book_name": "Ascendia Pharmaceutical Solutions",
  "txn_count": 407,
  "book_mismatch_severity": "high",
  “naics_code”: {integer},
  “naics_description”: {string},
  "daily_balances": {
  "2024-11-30": 1855880.01,
  

• Example of webhook response where Book mismatch severity is found (book.pacing.discrepancies_found and book.pacing.no_discrepancies_found events):

  JSON

  {
    "book_uuid": "1e03d9cd-7d84-4672-bc3d-c19cd19c",
    "book_mismatch_severity": "High",
    "uploaded_docs": [
      {
        "uuid": "db7fe37f-3fc8-4a0a-9066-94101a1",
        "name": "uploaded_doc.json",
        "created_ts": "2025-03-01 00:00:00",
        "overlapping_date_range": {
          "begin_date": "2025-01-15",
          "end_date": "2025-01-31"
        }
      }
    ],
    "event_name": "book.pacing.discrepancies_found",
    "notification_reason": "Discrepancies Found",
    "notification_type": "STATUS",
    "severity": "HIGH"
  }
  

• Example of the SMB exported Excel sheet where Book mismatch severity is found:

  

Reason codes for mismatches

To help underwriters understand why each mismatch was flagged, a new field, reason_code is now included for every mismatched transaction. These codes provide immediate clarity without requiring manual comparison of raw Plaid and PDF data. Example reason codes include:

• Posting Date Difference
• Non-Critical Transaction
• Critical Category Mismatch

To learn more, see the enriched transaction endpoint.

Related transaction linking

For mismatches caused by minor date discrepancies (posting date difference), Docs-to-Digital includes a related_txn_pk field that points to the likely match in the opposite data source. This gives underwriters the context they need to assess whether the mismatch is a simple posting lag or something more significant. To learn more, see the enriched transaction endpoint.

New Docs-to-Digital Excel sheet

To support customers who prefer working directly in spreadsheets, we’ve introduced a new Excel sheet within the SMB Excel export specifically for Docs-to-Digital insights. This Excel output makes it easy for even non-technical users to efficiently review reconciliation results without needing to interact with the API. This sheet includes:

• Book-Level Summary: Displays total mismatches and overall book mismatch severity.
• Comparative Metrics: Shows side-by-side comparisons of revenue, debt, NSFs, and MCAs between Plaid and bank statements.
• Critical Category Totals: Provides mismatch counts and dollar amounts broken down by risk type.
• Severity Summary Table: Summarizes the total net mismatched amount for each severity tier.
• Detailed Mismatches Table: Lists transaction details, severity levels, reason codes, tags, and source format.

To learn more, see the mismatched transactions via SMB Excel Export.

Detect - Business Verification

❗️

Feature deprecation - Middesk Integration

The following Business Verification APIs are now deprecated:

• Create Business Verification (POST)
• Retrieve Business Verification (GET)
• Vendor Authentication (PUT)

These endpoints are no longer supported and should not be used in new integrations.