CourseCrafter is a tool for building courses from your own material with an AI assistant (Roumpas & Xenos, 2024; Roumpas & Xenos, 2025). The design goal was to make an AI-driven workflow feel unintimidating: a drag-and-drop canvas as the obvious starting point, a prompt bar as the single conversational entry point, and a short, jargon-free action rail (My Files, Save, Start Over, Help, Exit) so nothing about “authoring with AI” feels like a wall of settings. The retro pixel treatment is a deliberate move to make an unfamiliar workflow read as approachable and low-stakes rather than technical.
CourseCrafter: a drag-and-drop canvas and single prompt bar for AI-assisted course building, with a deliberately playful pixel aesthetic to lower the barrier to entry.
tested with educators
I tested the design with ten practising educators on a high-fidelity Figma prototype (SUS 71.75). The finding that changed the project was about AI: most participants didn’t notice the feature at all, and once it was explained, every one of them supported it, on one condition: the educator has the final word on anything generated. The full study is on the educational technology page. (Roumpas & Xenos, 2025)
References
2025
-
Bridging the Gap Between Educators and Game-Based Learning: A Study of an AI-Assisted Course Creation Tool for Minecraft Education
In 2025 3rd Cognitive Models and Artificial Intelligence Conference (AICCONF), Jun 2025
This paper explores the involvement of artificial intelligence in the context of educational technology. Specifically, it presents findings from a study evaluating CourseCrafter, an AI-enhanced prototype designed to assist to the creation of Minecraft Education courses by educators, regardless of their coding expertise or familiarity with Minecraft’s controls. The study aimed to gather insights on user experiences, interface usability, and the effectiveness of AI integration within the tool. Ten educators, all of whom had some experience with Minecraft Education prior to the experiment, from different geographical areas, teaching grade and subject areas tested the prototype, completed a System Usability Scale (SUS) questionnaire, and attended a semi-structured interview. To ensure a controlled environment throughout the prototype testing, user personas were randomly handed out to participants, they were instructed to follow a detailed scenario, and alternative options would be explained with the Wizard of Oz method. The participants unanimously found the idea useful, agreed on the functionality of certain features, but shared varied concerns about the use of AI in education and the complexity of the platform. Additionally, feedback highlighted the importance of intuitive design, adequate instructional resources, and the potential impact of the tool on teaching and student engagement. These findings emphasize both the need for refinements to enhance user experience and maximize educational effectiveness but also the demand of such a tool in the industry.
2024
-
Design of an AI-Assisted Platform that will Facilitate Educators in Creating Minecraft Education Courses
In 2024 6th International Workshop on Artificial Intelligence and Education (WAIE), Sep 2024
The potential of Education Technology has reached a point where teachers cannot keep up with the latest trends, nor they can spend the time to get accustomed to innovative technologies that might become obsolete in a few years. With the rise in the popularity of Minecraft Education, an educational platform based on the engine of the world-famous video-game: Minecraft, as well as the benefits of gamification in classrooms for all student ages, it seems that an application that would help teachers create educational video games that are run on Minecraft Education, without any code and with artificial intelligence assistance, could be just the thing that they need to update their teaching techniques. We propose the creation of a platform like that, and we present its potential design, after being consulted by the available research that has been done to this day regarding instructional design, user interface and applications that utilize large language models.