Technology now helps financial institutions improve every part of the lending process. So Credit Unions may wonder how to compete with megabanks employing fleets of engineers building solutions. Enter Origence and its partnership with Informed.IQ.
During a recent webcast hosted by Origence, Justin Wickett, CEO and co-founder of Informed.IQ joined Bill Lynch, Origences VP of Partnerships. Wickett explained how Informed uses machine learning to classify, analyze, and extract data from documents for lenders. Lynch told the audience that Origence has partnered with Informed.IQ to create a document processing automation solution (DPA).
Wickett said that Informed works with numerous indirect auto lenders to help them fund dealers faster, improve compliance, and free up staff within their organizations.
The term “artificial intelligence” refers to the ability of computers to generally emulate human thought and perform real-world tasks, Wickett explained. Machine learning is a subset of AI and can be thought of as the technologies and algorithms that enable systems to identify patterns, make decisions, and improve themselves through experience and data.
AI presents several advantages to financial institutions, including allowing for more access to data throughout the organization and generating results with fewer human errors in real-time. He said AI is “on the rise” within consumer lending. Of financial institutions surveyed recently, 95% use AI or plan to do so in the near future.
The Current Process
Auto dealers are notorious for submitting 50-page PDFs containing multiple vehicle service contracts and other ancillary product contracts, Wickett continued, adding AI can break these apart in a consistent, accurate manner. It also can employ what is known as “fuzzy match logic.” In many cases, credit applications contain mismatches – for example, the name “Bob” used on one page and then “Robert” on another, abbreviations, and paystubs with the name of a subsidiary that is not the same as the company name on the check, and more.
“Today, credit union employees have to use Google or Bing to try to resolve these issues or make additional employment verification calls,” Wickett said. “AI harnesses industry data to make these matches.”
In a real-world example, Wickett pointed to a large indirect lender that sometimes took 30 days to get a dealer funded – in most cases because deal jackets were getting stuck in manual review.
“Delayed communication damages relationships,” he said. “With AI, this process is sped up exponentially. It can catch missing signatures and assess the legibility of various pages in seconds. DPA comes pre-trained right out of the box!”
Handling repetitive tasks
According to Wickett, consumer lenders frequently wrestle with when to leverage AI and when to use people. He posited that people do not go to work at a credit union because they enjoy performing repetitive tasks day in and day out.
“No one enjoys keying in information from contracts,” he offered. “There are many other important jobs, such as deepening relationships with members. Let’s let machines do what machines do well and free up staff for higher value jobs. I have heard repeatedly of staff being grateful for being freed from the repetitive task of checking signatures. Those tasks are error-prone and are actually a liability to the organization.”
With that said, Wickett noted AI works well with humans.
“You don’t want the AI guessing. Working with humans creates a feedback loop that retrains the AI and improves it over time.”
The Right Timing for Credit Unions
Informed got its start working with very large lenders because we wanted a broad array of data to train the AI models, but Wickett said the company’s partnership with Origence enables us to serve credit unions of all sizes.
“Credit unions are very unique in that they are open to collaborating with each other,” he said. “Even the smallest credit unions can tap into the data we’ve gathered. Credit unions can supercharge their lending process because the technology has been trained on millions of documents. Building a solution in-house takes a lot of effort and time. Even large financial institutions leverage Informed’s solutions – especially in fraud protection. There are 200,000 variations of fake paystub generators available online. Can a human memorize all of those types? AI can help sort through these in seconds.”
According to Wickett, the company is “democratizing” the data throughout the industry and bringing faster processing while improving compliance.
“We have already had success getting going with Origence credit unions and look forward to working with more going forward.”
A version of this article previously appeared on the Origence blog.
We are digital transformers of tedious manual tasks, bringing robotics process automation to financial services.