Delinkure harnesses machine learning to help mortgage lenders determine loans at risk
Fintech provider Speridian Technologies has launched a new tool that utilizes machine learning to predict loan defaults.
The new analytical technology, called Delinkure, enables lenders and servicers to manage portfolio risk by predicting delinquent accounts and providing resolution options.
"Due to the current economic situation, forecasts predict mortgage default rates will rise and be between 7% and 15%, with losses expected to exceed $200 billion ending 2021," said KP Hari, co-founder and managing partner at Speridian Technologies. "To efficiently handle this problem on a large scale, we leveraged the power of automated tasks and advancements in machine learning to help predict and prevent loans from becoming non-performing assets (NPAs)."
Delinkure includes two modules: default prediction and early resolution. Lenders can use the default prediction tool to categorize loan portfolio into high, medium, and low-risk watch-list groups based on risk scores.
The early resolution module analyzes various workout options via a workflow, which can be integrated with call centers for proper handling and easy resolution from loan modification agents.
"Delinkure uses up to date, real-time, external data which allows for specific disruptive events like pandemics and political elections. Our rapid deployment process or as-a-service offering with innovative pricing models including ‘pay-by-the-drink’ and gain-sharing, eliminate long-term projects and complicated agreements," Hari said.
