Name in original language
Uporaba umetne inteligence v poslovno inteligenčnem sistemu Skrinja
Initiative overview
Skrinja is operating in a secure GDPR-compliant government environment since 2020 and integrates major national datasets, including the Common Agricultural Policy (covering over 90,000 farms), Administrative Procedures (more than 1 million cases annually), Market and Health Inspections, Public-Sector Salaries (over 180,000 employees and 2,000 budget users), Public Procurement, the Central Personnel Registry and Public Funding Programmes. In 2026, Skrinja Open was launched to provide citizens, businesses, researchers and journalists with direct access to public procurement analytics, strengthening transparency, accountability and data-driven governance.
Skrinja as Business Intelligence (BI) platform (with data warehouse and UI ML supported predictive analytics) contributes to public value by enabling more transparent, accountable and evidence-based governance. It addresses the need for integrated and reliable data to support strategic decision-making, transforming administrative data into actionable insights for better public policies and more effective use of public resources.
Beyond transparency, Skrinja supports proactive policymaking through AI-powered predictive analytics. For example, machine-learning models help identify farms at risk of abandonment, enabling authorities to implement targeted measures before problems materialise. Combined with real-time monitoring of operational workloads and performance indicators, Skrinja allows public bodies to respond more quickly to emerging challenges, optimise resource allocation, and improve service delivery for citizens.
By transforming complex data into accessible and actionable information, Skrinja Open strengthens transparency and evidence-based governance (https://ejn.gov.si/skrinja.html, https://ejn.gov.si/skrinja/skrinjaeng.html). Skrinja also incorporates advanced AI and machine-learning capabilities, currently supporting the Common Agricultural Policy through predictive analytics covering more than 90,000 farms - over 80% of all farms in Slovenia. A predictive model based on XGBoost has been successfully implemented to identify farms at risk of abandonment, enabling early intervention and the development of targeted policy measures. The model analyses historical data and key farm characteristics to estimate risk levels, providing decisionmakers with valuable foresight through dedicated analytical dashboards. Building on this approach, similar predictive models are planned for other domains, including public procurement, administrative procedures, and inspection activities, further enhancing data-driven policymaking across the public sector.



























