For larger lenders, driving down cost, effort and time-to-close is crucial. When an underwriting team spends extra time clearing alerts, it slows down workflows and creates inefficiencies. Uncovering the highest fraud risk exposure using a state-of-the-art blended, predictive analytics approach provides lenders with the opportunity to gauge which loan applications require further due diligence.
The First American Data & Analytics AppIntelligence Score is an enterprise-level mortgage risk management solution that uses pattern recognition and advanced decisioning technologies, such as AI and machine learning, to deliver a highly accurate score and a precise picture of fraud and early payment default risk.
AppIntelligence Score was developed in response to large lenders’ needs for a predictive solution that would enable them to review a smaller sample of loans without increasing their risk exposure.
AppIntelligence Score streamlines the loan approval and purchase processes by quickly identifying and scoring loans with the highest and lowest risk. For example, loans with the highest fraud score, usually about 10% of total application volume, typically account for 50% or more of total fraud risk.
This targeting enables lenders to pinpoint and focus their reviews on the most at-risk loans, expediting decision-making and reducing operational costs. By targeting the scores in the top risk level, high-volume lenders can reduce reviews to as low as 5-10% of application volume.
While these proven, traditional systems work for most lenders, they are less efficient for high-volume lenders. AppIntelligence Score provides significant improvements in efficiency to enable the underwriting function to operate at a higher capacity, while targeting risk that can lead to losses or buyback requests.
The solution incorporates millions of data points in the First American fraud index and has been trained to identify patterns associated with different types of fraud schemes, including synthetic identity fraud.
Insights from decades of manual underwriter reviews and alert clearances have been fed into the model, so the solution can recognize which alerts were the most predictive of true risk. AppIntelligence Score has multiple sub-models to measure identity, employment, income and other types of fraud.
Lenders who use AppIntelligence Score will appreciate the operational efficiencies the solution delivers. Some lenders are also integrating the new AppIntelligence Score with traditional fraud alert models, like the First American FraudGuard model. With both, lenders get the advantages the AppIntelligence Score workflow solution offers and the insight into the score that FraudGuard can provide for auditing, alert-clearing and research.
When integrated with FraudGuard, users can use AppIntelligence Score to prioritize the kinds of alerts that they want to receive. Lenders can also override the score and review certain alerts and/or certain types of loans, regardless of their score.
The First American Data & Analytics solution provides lenders with significant improvements in efficiency, enabling the underwriting function to operate at a higher capacity, while targeting risk that can lead to losses or buyback requests.
Product Snapshot: The First American Data & Analytics AppIntelligence Score™ combines pattern recognition and artificial intelligence to quickly identify and score loans with the highest and lowest risk.
