2024-11-182024-11-182024-01-10PAMPOLHA, Pedro Augusto Duarte. Ferramentas de inteligência artificial da google para previsão de demanda empresarial. 2024. Trabalho de Conclusão de Curso (Bacharelado em Engenharia de Produção) – Universidade do Estado do Pará, Belém, 2024.https://bibliotecadigitaldetcc.uepa.br/handle/riuepa/513This study analyzes the significant contribution of Google's artificial intelligence (AI) tools, specifically BigQuery ML and Vertex AI, in the context of business demand forecasting. The relevance of this research lies in the growing importance of using these tools to improve the effectiveness and accuracy of forecasts that result in better business management strategies. The methodology adopted is a case study in which the AI tools BigQuery ML and Vertex AI are applied to demand forecasting. Initially, five sets of data were constructed from interactions with ChatGPT and Bing AI, each characterized by specific time series. This data, which was fundamental to the development of the research, was later integrated into the BigQuery platform for use in the AI tools in question. The methodological approach embraces the careful extraction of training tables from each data set, followed by the creation of prediction models and the subsequent generation of forecasts. The results were integrated into a visual platform to create comparative graphs, with evaluation metrics for analyzing accuracy and comparing the forecasts of the two artificial intelligences. The results indicate that BigQuery ML is more accurate at forecasting demand than Vertex AI in three of the five time series, except for the U-shaped series, where both tools performed similarly, and for the stationary series, where Vertex AI achieved more satisfactory results. In addition, BigQuery ML is faster at forecasting demand for all time series, both in training and forecasting. In terms of ease of use, Vertex AI proved to be more user-friendly when forecasting demand compared to BigQuery ML. We conclude that BigQuery ML outperforms Vertex AI in forecasting demand for time series with increasing, decreasing and inverted U-shaped trends, while both tools performed similarly for time series with U-shaped trends. This study contributes to understanding the performance of Google's AI tools in demand forecasting, providing practical insights for the business context.Língua PortuguesaAbertoArtificial intelligence toolsGoogleBusiness demand forecastingCompetitivenessData analysisTrabalho de Conclusão de Curso (TCC)