Detection of observations:

This was a work done in collaboration with the National Directorate of Public Procurement of Paraguay (DNCP) and the Data Science for Social GoodxGermany (DSSGxGermnay) program in the summer of 2022. In this project, a model was built to identify public tenders that are likely to present inconsistencies with the regulatory law, and provide auditors with explanations to better guide their review. The DNCP manually audits thousands of tenders per year and expects a higher volume in the future as the country's economy and population grow. Having a smaller, prioritized list of tenders with probable irregularities allows the DNCP to better leverage its resources and review more tenders with greater attention.