CASE STUDY: AI-enabled Auto Loan Document Processing


A national leading auto loan firm had a goal to streamline their processing and approval by organizing loan packages into a consistent document format using AI-enabled auto loan processing. This included classifying, separating, and reordering the pages of loan documents received from auto dealers so they are consistent.  Axis Technical Groups’ (ATG) goal was to allow them to quickly organize and prepare the loan documents in a repeatable fashion, resulting in time and cost savings over the manual handling of these complex documents.


The Client processes thousands of loan files each month, where all the document types are received from external parties as a single PDF. Their previous audit process required users to manually work through disorganized packets of images that could contain the various document types in no particular order, sometimes with packets missing pages or even entire documents. This process could take hours per loan and could not be performed by lower-cost junior team members because the loan underwriters required SMEs (Subject Matter Expert) to identify, correct, and sign off on the information contained in the document packet.


ATG’s AI auto loan solution automatically classifies, staples (virtually separates pages and builds document types back together), reorders and bookmarks the loan by the document type (i.e., Income documents, utility bills, identification documents like a driver’s license, etc.) into a consistent, easily digestible format, allowing the underwriters to locate the information more quickly and increase the speed of data entry.

Documents are first processed through an optical character recognition (“OCR”) process to convert the information into a computer “readable” format, after which the OCR’d data and images are classified and analyzed using Natural Language Processing (NLP), Machine Learning (ML), and Deep Learning technologies.

The required documents are then carefully packaged and returned as a PDF with a specific order to help streamline the data entry into the Loan Origination System (LOS). Integration between the Axis solution and the client’s LOS system helped to identify what the documents should look like, and also to alert the operator if any pages were missing from the loan package. The integrated approval workflow and interface bring attention to potential errors and allows the operator to move quickly through a package, only spending time addressing exception handling.

Next Steps

The next steps will be to extract the required data elements automatically and programmatically feed the LOS system. Data elements from the origination documents will be extracted, validated, and/or corrected with the values then passed to the client’s LOS for seamless ingestion. This will not only effectively eliminate manual data entry, but it will also allow lower-cost data entry staff to perform any required validation. This frees up the underwriters allowing them to focus on loan authorization and overall approval of the package. These combined savings in time and labor cost yield increased throughput, equating to tremendous savings per loan. Taking a conservative average of cost per loan multiplied by a number of loans brings the annual savings to potentially millions of dollars per year. Client executives stated: “The Axis Team is outstanding, to say the least, and we truly appreciate everything they have done to enable us to streamline our process! We are most pleased with the savings from Phase One alone and want to ensure that all future eligible loans are sent through the Axis process.”