POST-V: um sistema para análise preditiva de vendas de insumos agrícolas em uma cooperativa agropecuária

Detalhes bibliográficos
Ano de defesa: 2023
Autor(a) principal: Stöhlirck, Tiago lattes
Orientador(a): Hölbig, Carlos Amaral lattes
Banca de defesa: Não Informado pela instituição
Tipo de documento: Dissertação
Tipo de acesso: Acesso aberto
Idioma: por
Instituição de defesa: Universidade de Passo Fundo
Programa de Pós-Graduação: Programa de Pós-Graduação em Computação Aplicada
Departamento: Instituto de Ciências Exatas e Geociências – ICEG
País: Brasil
Palavras-chave em Português:
Área do conhecimento CNPq:
Link de acesso: http://tede.upf.br:8080/jspui/handle/tede/2541
Resumo: This work was developed in an agricultura! cooperative in the state of Rio Grande do Sul that has the concern and care to try to preserve and institutionalize all the technical know-how acquired over the years by its technical staff (agronomists, biologists and researchers). To this end, the technical area specializing in plant production prepares to each culture/crop a reference document, similar to a protocol, known as "technical positioning". This document describes the phases of the crop production cycle and the products and dosages per hectare (ha) that must be recommended by cooperative agronomists so that rural producers can maximize the productivity of their crops. Assuming that the recommended dosage of a product is 500ml/ha, soon it is concluded that for an application in a field of 80 hectares it will be necessary 40 liters of this product, and for two applications 80 liters will be needed. Since if there is the potential for selling this product per producer, and if the producer is part of a number of customers served by an agronomist, then it would also be possible to have the potential for agronomist, by the unit to which the agronomist is attached and, finally, of all the unities together, which would be the total sales potential of the cooperative. Due to these factors, in this work POST-V was developed, a solution that makes use of Data Science in arder to collect registration and historical data in the ERP system of this cooperative, to organize and analyze this data to generate knowledge about customers, to estimate the sales potential for each of them based on their cultivated area and the technical positioning recommendations, to make sales forecasts based on past harvests and to make this information available, through visualizations, in a that they could help the cooperative in its purchase planning, in the management of its inventaries, defining its goals, monitoring sales performance and in identifying sales opportunities with its members. The development of POST-V training allowed the cooperative to realize the benefits that could be added to the business by analyzing sales and customer data