Use this recipe to qualify mortgage loan applications by extracting income information found in supporting documentation such as scanned pay stubs and 1040 tax forms. You’ll turn hours of work into just minutes.
Loan originators must assess the ability of loan applicants to repay loans. Applicants provide evidence of their income by sharing copies of their pay stubs and tax returns. Typically, this information is scanned, faxed, or even photographed using a mobile phone. These unstructured file formats have traditionally required a human to manually review and extract the information, causing a bottleneck in turn-around times that ultimately leads to decreased customer satisfaction and lost revenue.
Impira AutoML technology allows lenders to corroborate images of paystubs against other data sources to allow them to quickly make credit decisions with a high level of confidence. Non-technical users can train a machine learning model to read the docs and images without ever writing any code. All it takes is just a few clicks in the Impira UI.
You can upload the PDFs, JPGs, PNG files directly into the Impira application (or set up a programmatic upload via our APIs). Upon receiving the files, Impira will use OCR and AutoML to classify the documents, identify the wages and income values, and automatically join all of this data together. With the information extracted and joined, lenders using Impira can even apply their own loan approval logic into our application - much the same way you would use formulas in a spreadsheet.
Data and calculations performed on that data in Impira can be programmatically sent to downstream systems within the lenders IT ecosystem. The result? Within seconds of receiving the files, lenders can turn-around a credit decision without any humans slowing down the process.