Using valued resources wisely is paramount to success in the
current mortgage environment. Leveraging robotic process automation (RPA) and artificial
intelligence (AI) means lenders and servicers can automate manual, repetitive
tasks and use the expertise of mortgage professionals where it will make the
LoanComplete by Fiserv uses RPA and AI to solve one of the most
common problems facing lenders and servicers who want to automate — the ability
to decode different kinds of documentation and use this information to automate
tasks typically performed by humans.
Using Intelligent Content Recognition
(ICR), which can classify both structured and unstructured data, LoanComplete
identify documentation from loan packages, detect and identify the final
version of those documents, and accurately extract loan data to automate data
The solution converts all pages of the document to text, allowing
it to then extract loan data information from borrower-supplied income
documentation, third-party documentation and other sources. The extracted data
is delivered to the business rules engine to automate manual steps that would
normally be performed by processors, underwriters and closers. The business
rules engine ensures data integrity, checks that data matches across all
required documentation and populates any missing or inaccurate data fields in
the LOS or Servicing System with the correct data.
LoanComplete uses loan data information to determine what
documents are required for a loan file based on conditional logic that
identifies whether documents are required or optional – one more way the LoanComplete
solution and AI lighten the workload for users, who would traditionally have to
do this themselves. The loan data information can include factors such as the
type of loan product, loan-to-value ratios, state specific documents and dollar
As documents are received and processed through the Intelligent
Content Recognition software, LoanComplete automatically attaches them to the
Loan Audit Record within its system. The documents can also be automatically
delivered and imported into a LOS.
This automation has resulted in significant efficiency gains for
Fiserv’s LoanComplete users, with processing cycle times reduced by up to 50%.
In addition, by automating the review process and having users review only the
exceptions, onboarding audit and quality control times have been reduced by up
to 70%. Now, lenders can increase quality control audits from a sample size of
loans to a full, 100% review throughout the Origination lifecycle.
“Implementing AI and RPA reduces the unproductive time mortgage
professionals used to have to spend reworking the loan by ensuring proper
documentation and data exist before moving onto the next stage of the
origination process,” said Gregg Lehman, Manager, Product Management, Financial
& Risk Management Solutions, Fiserv.
“Alerting users of problems with the loan early in the process allows issues to
be addressed quickly instead of appearing and needing to be fixed after the
loan has already been closed and funded.”
- Extracts income data from borrower-supplied
paystubs and W-2s, checking for accuracy across documents and eliminating the
need to stare-and-compare
- Delivers missing income data to the Loan
Origination System to eliminate the need for manual data entry
- Summarizes for both users and auditors a
complete and trackable loan record that demonstrates a record of compliance of
all documentation received and audited throughout the Origination process.
The ICR capability is just one part of the larger LoanComplete
solutions suite, which includes the Comparalytics business rules engine,
LoanTracker Case Manager and Delivery Manager solutions. Learn more at Fiserv.com/LoanComplete.