Search papers, labs, and topics across Lattice.
This paper evaluates a project-based master's course on AI-enabled systems, where students designed a movie recommendation system while navigating architectural challenges related to scalability and deployment. The study employs a mixed-methods approach, analyzing student submissions and questionnaire responses to identify integration difficulties and learning outcomes. Key findings reveal that students struggled with early architectural decisions and heterogeneous ML integration, highlighting a gap in software engineering expertise among learners.
Students developing AI systems face significant hurdles in architectural design, revealing a critical gap in current AI education.
Teaching Software Engineering for AI-enabled systems entails addressing the integration of AI components within full-scale software architectures under realistic constraints. While machine learning courses emphasize model development, students often lack experience in architectural design, deployment, and monitoring of AI-enabled systems. Empirical evaluations of such system-oriented AI courses remain limited. This paper reflects on the design and implementation of a project-based master's-level course titled AI Algorithms: Theory and Engineering, at the University of Bremen, in which students developed a movie recommendation system while making architectural design decisions to address challenges related to scalability, deployment, and evolving requirements. We conducted a mixed-methods study combining analyses of student submissions and questionnaire responses to investigate integration challenges, learning outcomes, and opportunities for improvement. Our results indicate persistent difficulties in early architectural decisions, heterogeneous ML integration, evolving requirements, and data management, largely due to uneven ML and software engineering expertise. From the educator's perspective, the course fostered system-level reasoning and strengthened awareness of data-centric ML practices in AI-enabled systems.
