Tame the Mortgage Refi Process with Intelligent Automation

February 25, 2021 / Commercial Banking, Intelligent Process Automation, Use Case

take the mortgage refi document mess

 

Sixteen times in 2020 mortgage rates fell to record lows, prompting an astounding 105% increase in mortgage refinance applications as compared to 2019. While that’s great news for homeowners as well the financial institutions writing all those mortgages, it also creates serious demand for intelligent automation in mortgage document processing. 

Consider that each mortgage refinance applications involves potentially dozens of documents: the loan application, appraisals, property valuations, lock-in agreement, real estate bills, pay stubs, tax returns and more. At most banks or lending institutions, humans have to read all these documents, pull out pertinent data and input it into downstream systems for processing. 

That approach needs an upgrade in an era where fintech startups are already disrupting all aspects of banking, including mortgages. MQube, for example, was founded in 2016 with the sole purpose of reinventing the mortgage process, including the use of artificial intelligence to make fast, informed decisions on credit worthiness. 

 

The trouble with templates and RPA 

Traditional banks and mortgage brokers can compete, if they likewise take advantage of technology including AI to help them deal with the document burden. Many have already created online portals that make it simple for applicants to upload relevant data. That’s a good start. 

The next step is to automate the processing of all those documents. Many financial institutions have dipped their toes into document automation waters by using robotic process automation (RPA) or templated approaches. This involves defining specific fields in a document from which data should be extracted. But given the variety of documents in question, banks quickly find that approach doesn’t get them too far. It’s simply not feasible to create a template for every type of document that may be involved in the mortgage refinancing process. 

 

Related Article: 3 Use cases for Intelligent Document Processing in Commercial banking

 

Hyperautomation for mortgage processing

What’s required is a solution with deeper AI roots that can deal with all sorts of documents, including unstructured content. The solution is a concept known as intelligent document processing, which falls under the “hyperautomation” umbrella in Gartner’s automation hierarchy. 

An effective intelligent document processing tool enables companies to automate much of the mortgage refinance application process. Rather than having a human read each document and extract data, the intelligent automation tool uses AI technology known as natural language processing (NLP) to perform that function – in a fraction of the time and with greater accuracy. 

The key is to ensure the automation tool you choose is built on a data platform that’s large enough to enable effective NLP. Indico’s Intelligent Process Automation platform, for example, is built on a database of some 500,000 million labeled data points. That gives it the intelligence to understand virtually any kind of document, including the context behind it. 

Indico then applies another AI technology, transfer learning, to make it easy to apply that massive database to any document processing task. It takes only about 200 “live” documents to train a model that will then be around 95% effective in automating mortgage document processing. 

 

Dramatic cuts in process cycle times 

That’s why Indico customers routinely are able to slash document processing cycle times by 85% – the kind of gains that enable them to effectively compete against the most nimble fintech.  To see for yourself how the Indico IPA platform can help get you out from under the mass of mortgage documents, arrange a free demo. Or, if you have any questions, feel free to contact us

Author
indico

Don't Miss a Post

Get our best content on Intelligent Process Automation sent to your inbox weekly.