Blend using artificial intelligence to improve mortgage processing

Register now

Blend, a provider of mortgage point of sale systems, is offering a new product that uses machine learning to streamline loan closings and analyze loan data quality.

The new technology, Blend Intelligence, analyzes data to produce the process improvements, according to an announcement the company made at the 2018 Digital Mortgage Conference.

"The engine is fed by vast quantities of high-quality data that continuously improve its performance over time," Grace Qi, Blend's product manager, said in an email.

Data analyzed by Blend Intelligence includes information from loans on Blend's platform. The company, which was founded in 2012, has collected more than $150 billion in loan applications in 2018. It has more than 120 customers.

Patterns that Blend's new product identifies in the data help improve the technology's ability to predict and automate mortgage lenders' workflow.

Functions that Blend Intelligence is automating include document requests and follow-up tasks for mortgage borrowers and third-party vendors. The technology also identifies and flags problematic loan data.

Blend Intelligence can be used to handle tasks such as automating a common lender request for a document like a pay stub. It also can help lenders handle rarer circumstances such as a need to ask a loan applicant to submit a profit and loss statement.

The San Francisco-based company designed the system to use predictive modeling to identify missing loan conditions in unusual situations as well as common ones so that lenders can address such issues earlier in the origination process than they otherwise would.

The private secondary market for nonstandard loans is growing.

For reprint and licensing requests for this article, click here.
Digital mortgages Machine learning Artificial intelligence Data quality Document management Underwriting Blend