Improving Fraud Management Operations

Energy & Infrastructure / South America / Fraud Detection

$2 BI

Nulla vitae elit libero

$4.3 MI

Fusce dapibus tellus

73%

Ipsum libero

Goal

Help a leading Latin American water utility increase detection of fraudulent customers, improve automation and efficiency across the fraud inspection scheduling process, and use data and analytics to improve prioritisation of customers, detect new fraud types and help front-line teams execute inspections more efficiently.

Approach

• Developed a robust data engineering pipeline infrastructure to perform hypothesis testing across data sources and automate the generation of daily inspection lists.

• Gained in-depth understanding of end-to-end fraud inspection operations.

• Reviewed past fraud inspection data, identifying key high fraud propensity features (e.g., irregular consumption, irregular payment behavior, neighborhood fraud activity levels).

• Developed a machine learning model that showed strong results in prioritizing inspections in simulations with historical data.

Interventions

• Designed a pilot for a specific concession to test model performance on daily prioritization and surfaced new fraud cases in the broader population before scaling the solution to multiple concessions and rolling it out to the second concession.

• Achieved an overall 1.8x uplift in fraud and anomalies caught, translating to an estimated $12M in additional revenue collected per year.

This uplift was driven by:

• Increased detection accuracies (up to 1.6x) and exposing new fraud types beyond existing rules.

• Improved operational efficiency and productivity (up to 2.3x).

• Value-based prioritization of customers with higher reconnection rates and recurring revenue.

• Trained and up-skilled the internal team to independently run the end-to-end solution and roll it out to other concessions across the country.