Release notes

In this release, we've introduced new Detect signals, new Analytics tags, and extended support for new Classify and Capture documents.

What's new?

Detect

Detect pay stub signals

We enhanced our Detect product with the following two new algorithmic signals on pay stubs:

Invalid date

This signal is triggered when the pay date on a pay stub falls on a U.S. bank holiday or weekend in the given year. It highlights instances where the listed pay dates do not fall on valid business days.

Impact on Authenticity score: 55 if no other signals are found
Level: Medium

Unreconciled gross pay

This signal is triggered when the gross pay does not match the sum of the current pay across all earnings categories, indicating potentially inaccurate pay stub calculations.

Impact on Authenticity score: 65 if no other signals are found
Level: Medium

To learn more about the tampering signals, see the Pay stub tampering signals.

Classify and Capture

New Classify and Capture supported documents

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

Property

Homeowners Association Statement

To learn more, see the All Supported Documents page.

Analytics

SMB lenders need full visibility into a merchant’s bank accounts, transfers, and payment activity to make well-informed underwriting decisions. By analyzing cash flow patterns, Ocrolus provides lenders with deeper insights into key aspects of merchant financial health and debt capacity.

In this context, we have introduced the following new features in the Analytics endpoints:

Book summary endpoint

The following two features are added to the Book summary endpoint:

Missing bank accounts

The presence of an unreconciled internal transfer suggests that borrowers did not submit their full financial picture to the lender during the underwriting process. The missing bank account metric is a new feature within the Book Summary Metrics endpoint to indicate which bank accounts were not submitted as part of the Book.

When a missing bank account is detected, Ocrolus will return the account number, whether the bank account type is Checking, Savings, or Unknown, and the total amounts transferred into and out of the missing account.

"missing_bank_accounts": [
    {
      "account_type" : "CHECKING" # or "SAVINGS" or "UNKNOWN",
      "account_number": "9066",
      "total_transfered_into_account": -10010.11,
      "total_transfered_from_account": 23234.23, 
    },
    {
      "account_type" : "CHECKING" # or "SAVINGS" or "UNKNOWN",
      "account_number": "8077",
      "total_transfered_into_account": -10010.11,
      "total_transfered_from_account": 23234.23,

Debt events

A new events key has been introduced to the Book Summary endpoint to enhance payment pattern analysis for Fintech loans and Merchant Cash Advances (MCAs). The following events are identified to reflect changes in payment behavior:

PAID IN FULL: Identifies when the total payment sum is greater than or equal to 102% of the total deposit sum.
STOP PAYMENT: Detects when payments stop entirely when a transaction contains a description including the term stop payment, or when payment amounts are significantly less than deposit amounts.
REDUCED PAYMENT: Flags a decrease in payment amount compared to the previous payment.
INCREASED PAYMENT: Flags an increase in payment amount compared to the previous payment.
PAYMENT FREQUENCY CHANGE: Captures changes in payment intervals compared to earlier patterns such as a change from weekly to daily payments.

"events": [
{
"event_type": "INCREASED PAYMENT",
"date": "2024-04-04"
},
{
"event_type": "PAID IN FULL",
"date": "2024-04-04"
},
{
"event_type": "STOP PAYMENT",
"date": "2024-04-04"
},
{
"event_type": "REDUCED PAYMENT",
"date": "2024-04-11"
},
{
"event_type": "INCREASED PAYMENT",
"date": "2024-04-11"
},
{
"event_type": "PAYMENT FREQUENCY CHANGE",
"date": "2024-05-16"
}
]

Enriched transactions endpoint

The following two features are added to the Enriched transactions endpoint:

Unreconciled internal transfers

Using this feature, Ocrolus now identifies whether transactions tagged as internal transfer are reconciled or unreconciled, using a new tag in the enriched transactions endpoint. An unreconciled internal transfer indicates that Ocrolus was not able to find both bank accounts involved in the internal transfer present in the Book. The transaction will contain the tag unreconciled_internal_transfer = TRUE.

In the case that Ocrolus was able to identify both bank accounts involved in the internal transfer within the Book, the transaction will contain the tag unreconciled_internal_transfer = FALSE. Additionally, it will return a reconciled_txn_pk for that transaction which corresponds to the transaction representing the other side of the internal transfer.

internal_transfer: true,
unreconciled_internal_transfer: true
reconciled_txn_pk: null

Identification of non-descript wires as debt inflows

We can now more accurately label non-descript wire transfers, such as transactions labeled WIRE TRANSFER without additional identifiers, as potential debt inflows. These transactions can be mapped to existing Fintech loans or MCA (Merchant Cash Advance) sources as applicable.

This enhancement to the Enriched Transactions endpoint ensures that select wire transfers identified as debt will receive an additional tag, such as Fintech Loan or Fintech MCA to indicate that the wire transfer represents a loan disbursement.