Check Fraud Detection Software
FraudOne is Softpro’s image-based, check fraud detection software, developed in collaboration with six leading US banks. With FraudOne, financial institutions can detect in real-time and with high certainty if a check is fraudulent or requires investigation prior to making payment. The automatic check fraud detection can be used for mobile, ATM or teller deposits of in-house checks. FraudOne also supports Day 1 and Day 2 back-office transactions.
The following chart describes an instant check verification process using remote deposit capture (RDC). (click image to enlarge)
The key to FraudOne’s reliable results is its decisioning technology – the Combined Risk Score (CRS) engine. The CRS takes into account a number of factors in order to determine if a check should be processed or investigated further. By combining data from a variety of sources and allowing the bank to set the rules and weightings, the false positives are minimized and analysts can focus their attention on suspicious transactions. In addition, all checks can be verified, not just those with high dollar value, further expanding your fraud protection.
Today, the top banks worldwide use FraudOne to curb their exposure to check fraud. In the US, our clients include Bank of America, Citi, KeyBank, Fifth Third Bank, Discover Financial, Synovus, and Banco Popular.
Watch a Real-Time Check Verification Demo
FraudOne maintains a database of reference signatures and check stock as well as additional account information. Check items are run through FraudOne in a real-time or batch mode to determine if the items are potential fraud. The pay/no-pay decision is based on a number of factors and scoring criteria set by the bank and powered by our unique Combined Risk Score (CRS) engine. FraudOne…
- Minimizes the false positive results; Signatures are cropped and cleaned to remove image artifacts, significantly improving the automatic verification process.
- Reduces operational costs associated with catching check fraud.
- Enables fraud analysts to focus on suspect transactions.
- Easily plugs into various fraud detection engines and other banking systems.
- Allows the bank to set the rules and weightings.
- Can be deployed in a Day 0 setting for real-time fraud detection of in-house checks. Also offers back-office Day 1 and Day 2 deployment options or a combination of all three.
- Was developed with the cooperation of the six leading US banks.
- Is currently used by 12 of the top 25 banks worldwide.
Combined Risk Score (CRS) Engine
Focusing solely on the signature or the check stock isn’t enough to catch the increasingly savvy fraudsters. A number of characteristics, managed by FraudOne’s CRS engine, allows you to more accurately identify if a check is fraudulent.
The CRS engine takes into account a number of weighted factors and leverages information from external systems in order to make a decision. It looks at whether or not the signature on file matches the one on the check; it identifies discrepancies in the check stock; and determines whether or not an item is a pre-authorized draft (PAD) and verifies if the payee is on a bank specified white or blacklist.
In addition, the CRS can accept data from external fraud engines and other bank systems. Prior to rendering its decision, the CRS applies the rules and weightings, established by you, to the myriad of data received.
If a check is deemed fraudulent, the item is placed into a visual verification queue for review by the bank’s fraud analysts. Analysts access the queue through the fraud analyst workstation which are client software applications that enable access to the back office system via a thin web browser, or a locally-installed client application.
FraudOne is comprised of two main modules, each one designed to handle a specific task of the check fraud verification process.
SignBase is the main FraudOne module. It is the reference database that holds the account information, including signature reference, check stock (reference image), user permissions as well as additional information such as scanned ID cards.
SignCheck is the database that holds the information needed to process the day’s work. The module uses the reference data stored in SignBase to perform the following tasks: automatic signature verification (ASV), check stock verification (APIA) and payee line recognition. SignCheck cross-references the payee to a customer-defined whitelist and blacklist. By leveraging the information in the whitelist, you can reduce the items flagged for analyst review and automatically make a no-pay decision for known fraudulent payees.
The analysts access SignCheck to review the items in the visual verification queue and make a payment decision. After the day’s final processing, SignCheck is cleared and prepared for the next day’s transaction.