Payment Decline Analysis Model & PBI Dashboard

Enabled full MBR & Financial Reporting Automation

Project Brief

A payments startup was struggling with elevated decline rates and, critically, had no systematic way to identify where to focus improvement efforts. Declines were visible in aggregate, but the team couldn't see which combinations of factors were driving them, so remediation was reactive and scattered rather than targeted at the patterns that mattered most.

I developed a statistical pattern-detection model in Python that surfaces high-impact decline patterns across multiple dimensions: PSP, issuer, geography, and transaction characteristics. This shifts the question from "why are declines high?" to "which patterns, fixed first, recover the most revenue?"

The analysis feeds a Power BI dashboard with full drill-down capability, letting the team move from a top-line view down to individual pattern segments to investigate root causes.

Note: synthetic data is shown to protect client information.