Summit

Federal Credit Fridays: Loan Monitoring—AI Predictive Insights Part 3!!!

Artificial intelligence is no longer a future concept—it’s reshaping how federal agencies monitor and manage their loan portfolios today.

In the latest episode of Federal Credit Fridays, host J. Anthony Curcio speaks with Katie Janik and Mike Rodriguez about how AI and machine learning are transforming federal credit oversight.

They explore:
🔹 Automated data extraction from reports and spreadsheets
🔹 AI-generated risk narratives that summarize borrower performance
🔹 Real-time dashboards and alerts that flag risks before they escalate
🔹 Best practices for integrating AI without losing the human touch

The result? Faster insights, smarter compliance, and a more proactive approach to risk management.

As agencies prepare for the next decade of modernization, one principle stands out: AI should empower people, not replace them.

🎧 Listen to the full episode: https://hubs.li/Q03N4ZrQ0

We're back with another where we dive into the evolving role of credit risk management in federal lending this week.

Anthony Curcio sits down with Andrew McCabe to discuss the impact of OMB Circular A-129, the role of Chief Risk Officers, and why embedding risk management early in the loan origination process is essential.

Key takeaways:
- Risk management should bridge origination and portfolio oversight.
- CROs succeed by balancing independence with collaboration.
- Strong communication is the foundation of effective risk governance.

🎧 Listen Here: https://hubs.li/Q03GbJr50

As the 2025 Joint Statistical Meetings continue this week, Albert Lee, Senior Founding Partner at Summit, will present tomorrow in the invited session Government Research on Predictive Analytics and Uncertainty Quantification.

His talk, based on a paper co-authored with Alan Salzberg of Salt Hill Statistical Consulting, Machine Learning Techniques Applied to Residential Home Price Estimations, will showcase how robust regression and bootstrapping improve the accuracy and reliability of Automated Valuation Models (AVMs). By addressing noisy data and outliers, their work highlights how predictive analytics can enhance transparency and support more data-driven decision-making in housing.

Additionally, Albert will participate in a roundtable on modernizing government statistical surveys. This discussion will explore how AI, cloud computing, and open-source tools improve the efficiency, reliability, and consistency of survey outputs. The roundtable will share federal agencies’ experiences implementing automated edits and imputations, emphasizing lessons learned in reducing manual effort, saving time, and improving data quality.