STEM with Generative AI: Fundamentals of Data Warehousing
K. M. Faisal, Andres Fortino · 2025
Data warehousing is a complex process that poses challenges for students, particularly in understanding star schema design, SQL coding, and ETL processes. To address these difficulties, we developed a ChatGPT-powered chatbot designed to guide users step-by-step in creating data warehouses, generating SQL code, and providing educational support for learning these concepts interactively.This work demonstrates the integration of multiple STEM disciplines by combining computer science (natural language processing, software engineering), mathematics (schema modeling, data relationships), and engineering principles (system design, iterative refinement) into a unified learning experience. The chatbot’s approach helps students understand how these disciplines interconnect in real-world applications, reinforcing the fundamental interdisciplinary nature of data engineering education.This proof of concept project aimed to test the chatbot’s effectiveness in enhancing student confidence, reducing SQL errors, and improving learning outcomes. We conducted structured trials where students interacted with the chatbot to design a star schema, create and query tables using SQL, and understand data warehousing principles.The results demonstrated an improvement in student confidence (110% increase) and improvement in SQL success rate (from 60% to 95%). Participants noted that the chatbot simplified the learning process by providing accurate SQL code and reducing the cognitive load associated with debugging and syntax management. However, minor challenges, such as incorrect table creation order in some cases, were identified and highlighted areas for refinement.The findings suggest that the chatbot can effectively serve as a scalable educational tool, offering both theoretical and practical benefits. It enables educators to streamline teaching processes and students to gain hands-on experience with immediate feedback. Future work will involve refining the chatbot’s error-handling capabilities, expanding its application to other domains such as advanced database concepts, and integrating visual aids for better comprehension.This project demonstrates the potential of AI-driven tools to transform technical education by reducing barriers to entry and enhancing learning efficiency. The main contribution is the development of an accessible, student-centered solution that connects theoretical knowledge to practical application in data warehousing.