AI in Lending Operations: Delivering Faster, More Accurate Outcomes

As lending markets continue to grow in size and complexity, so does the volume of data required to support investment decisions, portfolio management, operations and reporting. Critical information is often buried within loan notices, credit agreements and other unstructured documents, creating challenges for market participants who need timely, accurate insights.

While Artificial Intelligence (AI) is attracting significant attention across financial services, its real value in lending operations is not the technology itself. The value lies in helping clients access better information faster, with greater accuracy and less manual effort.

Turning Documents into Actionable Intelligence

Across both broadly syndicated loans and private credit, millions of operational documents are generated every year. These documents contain critical information on portfolio activity, borrower events, payment activity and amendments, but reviewing and processing them manually can be time-consuming and resource-intensive.

AI is helping to streamline this process by automatically reviewing and classifying loan agent notices based on their content. Within our Lending Solutions business, we are using AI to automatically extract 30+ data elements from agent notices, including the details of the agent sending the notice and the type of notice received (rollover, rate reset, interest payment, failure to pay or a revolver draw-down). AI then categorizes and digitizes a structured message to send downstream for straight-through processing (STP), while our experienced domain specialists validate outputs, review exceptions and ensure quality standards are maintained. This human-in-the-loop model helps deliver confidence and scale, with us seeing a reduction of 80% in the time taken to review and extract information from a loan document.

As a result, our team can support an astonishing volume of notices daily, more than two million on a quarter-end basis. We are now also able to complete cash reconciliation three days after the close of the quarter end, which is the highest level of efficiency ever achieved.

For clients, the benefit is straightforward: critical information can be identified, processed and delivered more quickly, helping teams respond faster to portfolio events and operational requirements.

Improving Straight-Through Processing

Another area where AI is delivering measurable value within our lending business is STP.

Many lending workflows depend on extracting key data points from agent notices and credit agreements. These documents often arrive in different formats and structures, creating challenges for traditional processing methods.

AI helps address this by identifying and extracting relevant information from unstructured content and transforming it into structured, usable data. Instead of directing users to where information might exist, AI-powered workflows can help deliver the specific data points required for downstream processes and reporting.

Human validation then helps ensure that data delivered to clients meets the high standards required for lending and investment workflows.

The result is a more efficient operating model, with faster delivery timelines, reduced manual intervention and more consistent outcomes across large volumes of documents.