Automating the Loan Review Process

United Acceptance Inc. PDF Download

United Acceptance Inc. (UAI) is a financial organization, founded in 1991, that provides capital and servicing solutions to automobile dealerships that have a "buy here, pay here" operation, finance companies, and independent finance companies. UAI has achieved major labor savings, eliminated a three-month processing backlog, improved corporate decision-making, and positioned itself for growth by automating its loan review process with a scanning and document capture solution from ibml.

The Challenge

UAI processes two types of loans:

  • Pre-funded loans: loans that UAI must review for potential purchase
  • Post-funded loans: loans that have been purchased

"We buy automobile loans from dealers across the United States. Once we have acquired a loan, we need to store the collateral package in a central location for our customer service representatives and other departments to view," explained UAI Information Technology Project Manager Laeeq Malik.

UAI's process of scanning and manually classifying loan documents created several challenges:

  • Burdensome document preparation: UAI's scanning and document capture solution required operators to insert separator sheets between each document to signify the start of a new document. Each document in a loan package also had to be hand stamped with an account number for manual tracking.
  • Slow scanner throughput: The document feeders on UAI's old scanners held only 100 pages. Scanner throughput ground to a halt as operators refilled the feeders. With the group of packages inside a proposal encompassing more than 1,000 pages, it was not uncommon for scanner operators to have to refill the document feeder multiple times for a single proposal. Making matter worse, UAI's old scanners jammed "pretty regularly," Malik said.
  • Software processing delays: After documents were scanned, it took up to an hour for the images to make it through UAI's network and to its documentclassification software. But the delays didn't end there: it took between four hours and five hours before UAI could output from its document-classification solution to its Microsoft SharePoint application.
  • Limited workflow visibility: UAI's scanning and document capture solution required operators to perform a number of manual steps to determine the status of a document.
  • Lost pages: With its old scanning and document capture solution, UAI ran the risk of losing scanned pages in memory if it waited longer than 5 minutes to save the work.

As a result, UAI had an up to 3½ month backlog of pre-funded loans that needed to be processed.

"When you acquire an auto loan mid-stream, the customer has a payment history with the dealer. Because of our processing backlog, we couldn't always answer a customer's questions regarding their payments. We would have to pull the file, scan it, and call the customer back," Malik said.

It is for these reasons that UAI's management began a search in 2014 for alternative solutions. After an exhaustive evaluation, UAI selected a scanning and document capture solution from ibml.

The Solution

After an evaluation of available solutions, UAI selected a solution that includes ibml's ImageTracDS 1085 scanner and SoftTrac Capture Suite software with SoftTrac Synergetics.

The ImageTracDS 1085 is a mid-volume production scanner that provides high-quality and efficient processing, and reduces document preparation costs, by allowing for the scanning of inter-mixed documents of varying thickness and size in a single pass. The ImageTracDS 1085 is equipped with two sort trays. This makes the device ideal for applications where checks, separator sheets or other sensitive documents must be returned to a customer or that require other special post-scan handling.

Synergetics, part of ibml's SoftTrac Capture Suite, is designed for organizations that need to handle multi-channel input streams and seek to automate document processing. Synergetics is an intelligent document recognition solution that classifies documents and extracts metadata automatically.

Malik said UAI selected ibml based on its impressive track record with U.S. banks, ibml's ability to handle high volumes of documents with tremendous accuracy for the U.S. Census, ibml's compelling video demonstration, and the results of a proof of concept that ibml performed in its Alabama offices.

The ibml solution has had a significant positive impact on UAI's daily operations.

During the capture process, ibml's ImageTracDS 1085 captures bi-tonal TIFF and color JPEG images of both the front and back of the documents. Once a batch has completed scanning, it is digitally transferred to ibml's Synergetics document-classification and meta-data extraction application. Synergetics classifies each document in a loan package and automatically determines whether six critical document types exist within the batch: cover page, retail installment contract (RIC), title, payment history, credit application, and bill of sale. If the six document types are found, the batch proceeds to extraction. If one or more of the document types are not identified, but the system finds a separator sheet noting the document is missing, the batch proceeds to extraction. If a document is not identified and no separator sheet is found, then the batch goes to verification by a UAI knowledge worker.

During extraction, Synergetics reads three fields – account number, first name (and middle name, if present), and last name – using optical character recognition (OCR) technology. The extracted data is used primarily for UAI's output file naming convention and indexing. After meta-data extraction is complete, PDF files are created from the images and moved to a specific location on UAI's SharePoint server.

The Benefits

The ibml solution provided UAI with a number of immediate operational benefits:

  • Streamlined document preparation: UAI no longer needs to insert separator sheets between each document. ibml's Synergetics document-classification solution automatically determines the start of a new document. UAI only needs to insert separator sheets in the event that a certain loan document type is missing (this assists with business rule verification later in the processing workflow). Eliminating the need for separator sheets between each document also reduces the money UAI spends on paper, and eliminates the time spent pulling separator sheets out of scanned packages. Additionally, UAI no longer needs to hand stamp an account number on each document for tracking. ibml's ImageTracDS 1085 prints the account number on each document in a loan package as part of the scanning process.
  • Faster scanning throughput: UAI replaced its four old scanners with two ImageDS 1085 production scanners. Additionally, UAI can now process multiple loan packages co-mingled. ibml's ImageTracDS 1085 reads a barcode on the cover page of individual loan packages to automatically separate each package into a batch. The cover page is also used to determine the correct account number that should be sprayed on each document in a loan package. Malik added that the ImageTracDS 1085 jams far less frequently than UAI's old scanners.
  • Reduced labor: The ImageTracDS 1085's throughput enabled UAI to reduce the staff required for scanning from 3 to 4 full-time equivalents (FTEs) per week to 1½ FTE.
  • Faster delivery of images to downstream applications: The transfer of images to ibml's Synergetics is a fraction of what it took with UAI's scanning and document capture solution.
  • Better pricing model: With its old document-classification solution, UAI paid a seat license, as well as a fee every time it surpassed 300,000 pages scanned. Malik said ibml's pricing model for its web-based solution is a better fit for UAI's growth. UAI now pays a single license fee, which allows UAI to have multiple people work on batches at the same time.
  • No processing backlog: UAI eliminated its backlog within one month of deploying the ibml solution. UAI now scans all of its loan documents within 24 hours of their purchase. Eliminating its backlog has allowed UAI to start scanning approximately 6.3 million pages from legacy accounts.

The ibml solution has worked so well that UAI is planning to scan all of its prefunding documents to allow its underwriting and quality assurance departments to simultaneously access loan documents.

"This will help us make quicker decisions on loans," Malik explained. UAI may also expand the solution to compare documents for pre-funded and funded loans to help identify discrepancies.

The Bottom Line

"This was the most positive experience UAI has ever had in deploying a new system and hitting a go-live date," Malik said. "We were particularly impressed with ibml's straightforward training, the ease-of-use of their system, and their excellent communications and project management."