Income Automation

Using OCR technology and deep learning models to classify documents and define the extracted data. I delivered a rule driven solution to automate the underwriting of income documents. 

Taking the lead as one of the first in the industry, if not the first, I led the team to develop an internal model using payroll direct deposits to calculate qualifying income in 2018

Bank Statements are challenging. Underwriters review all transactions within an account, so there is a lot of information and data to take into consideration for automation limiting the potential automation rate and efficiency gains. But I found a way to fully automate 60% of bank statement conditions. 

Automated the collection of income and asset documentation, using APIs with vendors, reducing the need to request documents from the customer. 80% auto-order of income and employment verification.

On-going development and management of internal calculators and verification tools. 

Deployed our deep learning model to extract over 600 data fields across 14 documents to support the underwriting process.