From simplifying loan processing to streamlining customer onboarding, intelligent document processing (IDP) captures, extracts and processes data for banks and other financial institutions.
To succeed with IDP, it’s critical to start with a clear business goal. Bank Automation News reached out to experts for best practices.
Banks need to know which documents they’re targeting for processing and how the technology will be used to support the business, Cheryl Chiodi, a financial services solutions manager and natural language processing expert at IDP vendor ABBYY told BAN.
“Banks need to have a clear understanding of what they want to achieve and have them actionable with business goals,” she said.
Chiodi pointed to the federal Paycheck Protection Plan (PPP) loan process, for which banks coupled IDP with robotic processing automation (RPA) to handle the U.S. Small Business Administration (SBA) loans in a timely manner. One of the banks ABBYY worked with was the $1.1 billion First Home Bank, a commercial bank based in Saint Petersburg, Fla.
“We were able to ingest, classify and extract data from those documents and we worked heavily with one of our RPA vendors to complete the process and expedite that,” Chiodi said. “The added complexity with PPP in particular, is that there’s a need for forgiveness and forbearance as well, so there needed to be an audit trail of all of those documents that we were helping First Home Bank to ingest, and extract and classify.”
Specifically, IDP helped move information into the E-Tran portal used by the SBA to process loans. Without IDP, it would have been a manual process, potentially filled with errors, Chiodi added.
Banks can use intelligent automation to transform their processes, Grant Vickers, of IDP solution provider WorkFusion, told BAN.
“The first thing is prioritizing IDP capabilities for very specific business problems that FIs want to solve for,” Vickers said. “We see the most important focus to be on the long tail end of complex documents and communications across the customer lifecycle.”
IDP solutions will typically use optical character recognition technology to digitalize documents. These solutions “cater to a semi-structured world,” said Anil Vijayan, a research analyst with the Everest Group. This makes IDP ideal for forms such as invoices.
For example, the $176 billion Citizens Bank used IDP to automate importing data from the financial statements of commercial borrowers — a process that previously relied on manual keying, according to Vinay Jha, chief data officer at Citizens.
More banks are turning to IDP to automate, said Sam Bobley, co-founder and CEO of IDP provider Ocrolus.
“I think process automation in general, is very hot right now, and particularly unlocking data that’s been trapped in documents and finding unique ways to turn it into digital data that can live for a life cycle,” he said. “It’s been resonating extremely well in the market and we think that the banks are eager to learn from some of the FinTech lenders who have those modern digital processes.”
The ability to extract data from semi-structured forms makes IDP a flexible solution for banks. Here are other IDP use cases that experts shared with BAN:
1. Customer onboarding
IDP can automate the customer onboarding process to improve the overall experience and reduce data quality issues in the customer profile.
IDP can streamline know-your-customer (KYC) processes and automate the document intake process to ensure compliance with anti-money laundering (AML) regulations using IDP.
3. Loan lifecycle
IDP can accelerate the loan lifecycle, reducing complexity and enabling faster funding from servicing to origination, underwriting, document capture and closing.
IDP can be used to improve trade service quality levels, expand trade operations, and ensure compliance to regulatory bodies to avoid fraud and fines.
5. Automation retail bank processes
Automating the application processes for credit, small business, personal, mortgage and auto loans, and account services are all potential use cases for IDP.
“There are just hundreds of thousands of contracts that are referring to the London Interbank Offered Rate (LIBOR) in banks,” Chiodi said. “Natural language processing will allow the scanning of all of these contracts in these various financial documents that refer to LIBOR, and be able to detect and, and then take action upon where that term is being used.”
The Bank Automation News webinar on automation technology for exceptional bank cybersecurity and ID verification takes place on Thursday, Sept. 9, at 11:30 a.m. ET. Register here. Attendees will be able to ask questions via chat.