Real-time Mortgage Fraud Prevention Through Automation at Fannie Mae

Customer Challenge

A major mortgage provider faced a critical challenge: staying ahead of sophisticated mortgage fraud. While they diligently responded to tips from customers, law enforcement, and concerned citizens about suspicious loans and anomalous activity, their investigative process was severely hampered. Investigators were mired in manual application checks and time-consuming dataset analysis, attempting to discern current fraud trends. This analog approach wasn't just inefficient; it created a significant risk exposure, as the sheer volume and complexity of potential fraud outpaced their human capacity. The client urgently needed to automate fraud assessment, empowering them to rapidly identify, thoroughly investigate, and proactively prevent mortgage fraud using cutting-edge technology.

Navitas Solution

To tackle this, the client partnered with Navitas to implement a real-time fraud prevention solution. A dedicated nine-member team, including developers and an agile coach, embarked on a three-phased development approach. The initial MVP delivered a cloud-native application with a self-service portal for consumers to report tips. They also deployed and customized an ML-based COTS solution with classification models to detect prioritized loan fraud scenarios using limited internal data, resulting in a significant 200% improvement in internal identification. The second phase focused on integrating external data sources and building custom ML models with automated CI/CD pipelines, streamlining development and deployment through an ML Foundation accelerator built on open-source technologies. In the final phase, Navitas leveraged entity resolution techniques to create a model that connected various data aspects, represented as Tuples and Dossiers. Throughout the entire process, agile methodologies, including test-driven development (TDD) and continuous integration and continuous delivery (CI/CD), were rigorously applied to accelerate the delivery of high-quality code.

Results

By partnering with Navitas, the mortgage provider successfully transitioned from a reactive, manual, and inefficient fraud detection process to a proactive, automated, and highly effective real-time fraud prevention system. The initial MVP alone delivered a staggering 200% improvement in internal identification of fraudulent activities, dramatically accelerating their ability to spot suspicious loans. No longer mired in manual checks and time-consuming data analysis, investigators could now leverage cutting-edge ML models and integrated external data sources, significantly reducing the risk exposure that stemmed from their previous analog approach. The implementation of automated CI/CD pipelines and entity resolution techniques ensured continuous improvement and a holistic view of potential fraud, empowering the client to not only rapidly identify and thoroughly investigate but also proactively prevent mortgage fraud with unprecedented efficiency and accuracy.

Solve Boldly with AI Get in Touch