Mapeamento de processos (BPM) para prevenção de fraudes em compras públicas: uma aplicação com o Sistema de Alertas Alice

Abstract

This article investigates the effectiveness of ALICE (Public Procurement, Contracts, and Notices Analyzer), an Artificial Intelligence (AI) system developed by the Office of the Comptroller General of Brazil, aimed at detecting risks in public procurement, based on a case study of an alert issued to the Federal University of Amazonas (UFAM). A qualitative descriptive approach is adopted, employing documentary analysis and process mapping through Business Process Management (BPM), modeling both as-is and to-be workflows. The results indicate that the current process, although ensuring formalization and traceability, is excessively bureaucratic, fragmented, and lacks prioritization mechanisms, which undermines the effectiveness of control. The proposed model preserves formal requirements while introducing immediate communication with the bidding committee, preventive suspension of procedures, decision-making centralization within internal auditing, and a continuous improvement cycle. From a theoretical perspective, the study integrates BPM and Actor-Network Theory, treating ALICE as a sociotechnical actor.