Why Pay for an AI Tutor When You Can Design It? Student-Created AI Agents as Co-Designed Instructional Tools for a Large Software Design and Engineering Course

Abstract

This paper presents insights into how undergraduate students in a large software design and engineering (SWE) course engage in the design of conversational intelligent agents (i.e., AI agents—AI stands for artificial intelligence) to support their learning process and build their technical skills. Students were asked to design and implement an AI agent on Dialogflow CX to help explain three software design patterns to fellow students. The assignment proposed they built an AI agent that could be preferred over other popular large language models (LLMs) when studying those three software design patterns. We report on students’ perceptions of learning, performance, the time they spent on the assignment (i.e., dedication), and their overall insights into what they enjoyed and disliked. We analyze students’ perceptions and insights based on their backgrounds and situational factors, reflecting on how student-created artifacts promoting the design and use of AI agents can help support learners from varied backgrounds in large computing classroom formats.