This case study presents the results of a project to develop a machine-learning supported risk assessment tool with the Special Investigation Service of the Republic of Lithuania (STT). The initiative was implemented in close collaboration with STT, the OECD and the Government Transparency Institute, with financial support from the European Commission through the Technical Support Instrument. The project aimed to explore how advanced data analytics and artificial intelligence can strengthen risk-based oversight of public spending by systematically identifying patterns associated with fraud and corruption. By integrating multiple administrative data sources and applying a transparent and explainable modelling framework, the model seeks to enhance the efficiency and effectiveness of investigative prioritisation while maintaining institutional accountability and professional judgement.