Loan Portfolio Analytics

A regional lending institution witnessed a spike in Non-Performing Assets (NPAs) but lacked the granular data to identify whether the cause was branch-specific, sector-specific, or due to external economic factors. Their existing reports were flat and failed to surface geographic or demographic risk clusters.

Field officers were struggling to prioritize collections, which led to further deterioration of the loan book quality. Management needed a way to visualize risk before it materialized into default.

I implemented a comprehensive Loan Portfolio Analytics platform using Tableau and PostgreSQL. The centerpiece was a geo-spatial heatmap that allowed management to drill down from regional trends to specific branch performance and individual loan accounts.

I integrated early-warning signals (EWS) based on payment history and external credit scores, allowing for proactive intervention. The dashboard also featured a 'Collection Efficiency Index' (CEI) to gamify and track field officer performance.

I also developed a 'Credit Score Correlation' engine that compared internal applicant data against historical defaults, helping the risk department refine its underwriting criteria for high-risk regions.