Diverse Claire: An AI-Powered UDL Approach Bridging Educational Experience Gaps in CS1

Providing equitable and accessible learning materials in computer science can be both challenging and time consuming, particularly for novice teachers or those uncertain about how to implement accessibility effectively. This issue is magnified by the lack of clear, actionable directions and training available for educators to put into practice. Diverse Claire is a tool that addresses this issue by providing teachers with structured, practical feedback on their teaching materials through the use of simulation. It utilises GenAI to analyse the material for accessibility and the Universal Design for Learning (UDL) framework as a guideline for analysis. This demo provides a walkthrough of how Diverse Claire analyses teaching materials, highlights potential accessibility barriers, and offers suggestions for improvements. Moreover, the system allows educators to understand how a computer science student with neurodiverse needs might respond to their course materials, by using a large language model to simulate responses to various test questions. These simulations are grounded in instructions drawn from research on neurodiverse learner experiences. The questions include short answer questions and coding exercises that are aligned with the pedagogical framework, Bloom's taxonomy, to target varying levels of understanding. In doing so, Diverse Claire supports CS educators in creating a more inclusive computer science learning environment that better meets the needs of all students.