While generative AI offers significant opportunities for learning, many Gen-AI applications tend to overpromise, particularly in the context of math and problem-solving courses. Some platforms provide inaccurate solutions, others offer correct answers without enough explanations, and some deliver responses that are not aligned with the student's level of understanding.  

The Math Learning Specialist and Peer Tutors at Concordia explored the development of a Gen-AI prototype that aims to deepen students’ understanding of a math course content and the importance of tailoring it to Concordia students. This presentation will focus on the design considerations in the development of this prototype.