Predictive analytics for efficient enforcement
Social Security Institute improves revenue through intelligent case selection
Challenge
Uruguay’s Social Security Institute needed to optimize its taxpayer case selection process by incorporating objective, data-driven criteria. The challenge involved analyzing and cross-referencing large volumes of information from multiple internal and external sources to prioritize cases with higher probability of irregularities, fraud, or evasion, improving inspection effectiveness and resource allocation.
Solution
SISCA, the Case Selection System, was developed as an automated and comprehensive solution to identify, analyze, determine, and prioritize cases for enforcement based on taxpayer behavior.
The platform integrates multiple data sources, applies statistical and predictive models, and provides analytical support for decision-making. It also enables continuous monitoring of enforcement outcomes to refine and strengthen future strategies.
360
Faster than the original process (5 days vs 20 minutes)
50
Increase in effectiveness
90
Selection effectiveness achieved
Benefits
The value behind the solution
More precise and effective case selection
Predictive models significantly improved the quality of enforcement targeting.
Better resource allocation
Behavior and propensity-based prioritization focused efforts on higher-impact cases.
Integrated taxpayer view
Combining multiple internal and external data sources enabled detection of behavioral patterns and irregularities.
Testimony
In executive reports, it is a fundamental management and decision-support tool.
CPA Liliana Mella
Technical Director, BPS ATYR
Our methodology
The project followed a structured and collaborative approach, integrating business and technology teams from BPS alongside Quanam specialists. From the outset, strategic objectives, prioritization criteria, and effectiveness metrics were aligned to ensure measurable impact.
An integrated data architecture was designed to consolidate multiple internal and external sources into a unified repository, ensuring consistency, traceability, and scalability. On this foundation, analytical and predictive models were developed to classify taxpayers, identify behavioral patterns, and prioritize cases with higher probability of irregularities.
The solution was integrated with existing systems and complemented with dashboards and monitoring tools that enable continuous performance evaluation and strategy adjustment. The project included process definition, controlled testing, and knowledge transfer to ensure technical autonomy and long-term sustainability.
Implementation was carried out in phases, starting with the Enforcement area and later extending to Collections Management, consolidating a cross-functional analytics platform to support data-driven decision-making.
Ready to take
the next step?
More success stories

Hyperautomation to modernize import approval processes
Ministry of Industry, Energy and Mining adopts RPA and AI to speed up validations and improve operational efficiency

Digital transformation in construction permit management
City Hall (IMM) implemented Generative AI to streamline procedures and improve services for professionals and citizens

