QuickML Creator: uma ferramenta de low code independente de plataforma para desenvolvimento de aplicações web portadas de modelos IA

• Ano de defesa: 2025
• Autor(a) principal: Rodrigues, Thanize Assunção
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Santa Maria; Brasil; UFSM; Programa de Pós-Graduação em Ciência da Computação; Centro de Tecnologia
• Palavras-chave em Português: Aplicação-web; Deploy; Frameworks; Ferramentas de automatização; Web-application; Automation tools; CNPQ::CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO
• Link de acesso: http://repositorio.ufsm.br/handle/1/36983

Technological advancement has optimized communication and the dissemination of information, with machine learning standing out as an essential technology for automating the construction and deployment of predictive models. However, deploying these models in real-world environments faces challenges, particularly regarding interoperability across platforms. Frameworks such as Streamlit, Dash, and Gradio simplify this process, but it is necessary to compare them to select the most suitable option. This study conducted such a comparison, evaluating performance, usability, and ease of deployment. Based on the results, the tool QuickML Creator was developed, which automates the development of applications integrated with trained models, facilitating their use for users with limited technical knowledge. The tool’s validation, carried out with 15 participants, indicated high usability, efficiency, and reliability. Thus, this work contributes to the popularization of machine learning usage, proposing solutions that combine agility, accessibility, and technical quality. The dissertation is organized into six chapters covering introduction, theoretical background, related work, methodology, results, and final considerations.