Cotação de materiais na construção civil: análise comparativa entre o método tradicional e sistemas automatizados com inteligência artificial

Abstract

Supply chain management in civil engineering, specifically in material quoting, is historically marked by analog routines and heavy human effort. This study quantitatively evaluates the efficiency of the quoting process in the private sector, proposing a technological solution integrated with generative artificial intelligence. The research is applied and experimental, grounded in Design Science Research (DSR), using a sample of sixteen regional suppliers and a standardized list of fifty materials. For the validation of the artifact, a direct comparison was made between the operational metrics of the conventional manual process and the performance of the developed automated system. Results demonstrated that the manual workflow presents a 25% attrition rate, an average response time of over six hours, and requires 3.3 hours of active effort for data tabulation, which generated technical inconsistencies in all valid proposals received. Conversely, the automated system executed the matrix data crossing in parallel in just seven seconds, ensuring absolute technical adherence with zero percent omissions or specification deviations. The usability evaluation indicated a minimal learning curve and significant perception of practical value, achieving a Net Promoter Score (NPS) of 96, positioned in the market excellence zone. In conclusion, the digital transition in the procurement sector optimizes pre-executive planning time and mitigates the invisible costs of bureaucratic rework, converting unstructured data into instantaneous strategic decisions.