AUTOMATED RECOGNITION AND CLASSIFICATION OF LOAN DOCUMENTS THROUGH LUMINATE™

Know More

CLIENT

A major national bank and a top lender in the United States was using stare and compare manual processes to classify, extract and upload correspondent mortgage lending loan packages, resulting in increased processing efficiencies and cost.

SOLUTION

Mphasis Digital Risk implemented Luminate™, an automated mortgage document processing platform that helped recognize over 500 common mortgage loan documents, including FHA, VA, and USDA forms. Leveraging ADR and OCR, Luminate™ reviewed the lender’s newly submitted, complete correspondent loan packages (comprised of PDF documents) of approximately 4,000 loans/month and performed three distinct processes:

• Classification: Determining the type of document (examples: application, appraisal, credit report)

• Versioning: Determining version of document (examples: initial/final application, initial/revised/final closing disclosure)

• Data Extraction: OCR (with QC assist) of client specified fields from document (examples: borrower name, appraised value, front/back end DU ratios)

Automated reports were generated from output of data extraction, configured to specifications, and missing document report listing based on identification and versioning.

BENEFITS

Our solution catalysed the following benefits for the client:

• Automated recognition and classification of each loan document in the package, eliminating manual processes.

• Identified client specific fields from each unique loan document to extract key data, saving time and cost.