Generative AI in building application for certification training

Abstract

This thesis explores how generative artificial intelligence (AI) can be used to create a web application that helps users prepare for official cloud certification exams. The project tested whether modern AI models, such as OpenAI’s GPT-4 and Anthropic’s Claude 3.5 Sonnet, can generate realistic and structured practice questions. The application was developed using technologies like Azure Functions, React, and Stripe for subscriptions. It uses prompt engineering to guide the AI in generating multiple-choice questions based on different certifications. By combining models, the app improved variety and reduced repetition in the questions. The results show that generative AI can support certification training by producing useful questions, but it still needs human oversight to ensure accuracy and formatting. More complex question types, such as case studies, remain difficult to generate with current models. The rapid development of newer AI models suggests strong future potential. Overall, this thesis demonstrates a practical way to use generative AI in education, while also highlighting areas for improvement in quality control and automation.