The mortgage origination and servicing processes are dependent on massive amounts of unstructured documents (e.g. pay stubs, W-2s, bank statements, tax returns, appraisals, etc.), which require extracting data from hundreds of pages content.  Documents must be verified to ensure completeness and traceability back to the System of Record (SOR).  Further, originators must create timely and accurate regulatory submissions (e.g. Uniform Closing Dataset) if selling the loans to Freddie or Fannie, and the data must be of high quality for broader reporting purposes (C-CAR, HMDA demographic data for the Uniform Residential Loan Application) and audit/control function activities (i.e. Internal, RESPA, and TRID).


  • Automatically attributed documents and extracted key data elements
  • Identified and mapped each document type in the loan file to bring structure to this mass of unstructured content
  • Highlighted potential changes in key data from the underlying loan documents (original source) versus the SOR and created accurate and timely reporting data sets to yield a more robust view
  • Developed transformational loan origination and servicing processes by integrating with third-party workflow and RPA/bots systems that  receive and case manage Pendo’s structured data sets

Value Created:

Transformed our client’s mortgage loan origination and servicing operational processes from a 100% manual, offshore model to a 98%+ automated solution providing substantive operational lift and creating auditable and monetizable data lineage and transparency for all of their lending documents.  The Pendo Solution enabled our client to offer new products and services (e.g. LIBOR assessment, C-CAR, CECL, Digital Compliance) by monetizing the insights and analytics trapped in the corpus of unstructured documents and surface meaningful marketing / BD opportunities.