CHATBOTS INTELIGENTES: UMA ANÁLISE COMPARATIVA DE TECNOLOGIAS E SEUS IMPACTOS NA EXPERIÊNCIA DO USUÁRIO EM PLATAFORMAS DIGITAIS

Abstract

This paper performs a comparative analysis of functional performance and user experience in generative AI platforms (LLMs) specifically Google Gemini, OpenAI ChatGPT, and Microsoft Copilot applied as productivity tools for students, professionals, and technology enthusiasts. A mixed-method approach was employed, combining a literature review with researcher-controlled practical tests to evaluate metrics regarding response time, factual accuracy, code generation usability, conversational naturalness, and user satisfaction. Interactions were recorded and assessed both qualitatively and quantitatively using rubrics based on Software Engineering and User Experience concepts. The obtained data revealed that, while there is parity in factual accuracy (Scenario A) and conversational fluency (Scenario C), critical divergence is associated with usability (Scenario B) and overall reliability. Google Gemini demonstrated superiority as the only platform to adhere to Software Engineering best practices, while OpenAI ChatGPT presented a critical technical failure, compromising its reliability. It is concluded that the effectiveness of these tools for the professional user is determined not only by response accuracy but also by efficiency, reliability, and, crucially, adherence to engineering best practices.