MiMaizey
WHAT IS MIMAIZEY?
MiMaizey is one of the University of Michigan’s official AI services, designed to help students and faculty access campus information, resources, and workflows through a conversational interface. Unlike GenAI tools like ChatGPT, MiMaizey is built specifically for the U-M community, offering secure, university-trained responses and deeper integration with campus systems. This redesign project was completed in collaboration with a client from the University’s AI department, who challenged our class to reimagine how these tools could better engage users and highlight their unique value. My work focused on enhancing MiMaizey’s usability, visibility, and overall appeal to make it a more seamless part of students’ academic experience.
ROLE
UX researcher, UX/UI designer
TEAM
Independent Class Project
TOOLS
Figma
THE PROBLEM
Students often overlook U-M AI tools because they don’t understand what these platforms do or how they fit into their academic lives.
At the start of the semester, our class met with Michael Hess, the lead of ITS Services, who emphasized that platforms like UM-GPT and MiMaizey were created to set a standard for Academic AI, one that goes beyond generic tools like ChatGPT by supporting student needs, campus workflows, and personalized learning. Yet, even with this vision, students don’t understand what these tools can do, why they’re different, or how they fit into their academic routines. This disconnect between institutional goals and real student experience reveals the central problem my redesign aims to solve.
THE SOLUTION
To bridge the gap between institutional goals and student needs, I redesigned MiMaizey to clearly communicate its value, streamline student workflows, and highlight capabilities unique to Michigan’s AI ecosystem. By adding guided actions, improving source integration, and creating a clearer, more engaging landing page, the new design makes MiMaizey feel intuitive and purpose-built for students.
USER RESEARCH
Where do students struggle to understand or use MiMaizey effectively?
Although this project centered primarily on UX Design, I conducted light foundational research to better understand how students currently perceive and use Michigan’s AI tools.
While MiMaizey is available to students, faculty, and staff, I prioritized students as the primary audience since they represent the largest user base and were the group most frequently cited by ITS as lacking awareness and engagement.
To inform early ideation, I ran a small round of informal usability tests and feedback sessions with students who had briefly interacted with the existing MiMaizey interface. A few consistent themes emerged:
After receiving feedback and refining my mid-fidelity wireframes, I developed the final design: bringing clarity, stronger guidance, and a more student-centered experience to MiMaizey.
Optional Chat History
Users can choose whether to save past conversations, giving them more control over their privacy and continuity.
This supports autonomy and reduces privacy-related anxiety, improving student trust and comfort.
IDEATION
Translating the core problems I identified into student-centered design opportunities for MiMaizey.
This is where I began turning insights from Michael Hess and the students I spoke with into early feature concepts.
I then used low-fidelity sketches to quickly explore layout ideas and potential workflows with these key features in mind.
From there, I translated these rough sketches into mid-fidelity wireframes in Figma, refining the structure and core interactions to shape a clearer direction for the final design.
FINAL DESIGN
App Integrations
MiMaizey connects with key campus apps to deliver more personalized, context-aware support.
Centralizing tools reduces cognitive load and supports seamless workflows.
Revamped Landing Page
A clearer, more engaging landing page introduces key features and highlights all of U-M’s available AI services.
A well-structured, transparent entry point reduces confusion, and clearly outlining use cases and elevating related services reduces ambiguity.
Model Switching
Seamlessly switch between U-M AI models within the chat to get the responses best suited to their task.
Offering choice accommodates diverse problem-solving preferences and promotes other U-M services.
Smart File Retrieval
The assistant can pull course materials, documents, and resources directly for the student when needed.
This minimizes search friction, supporting users with executive function challenges and busy schedules.
REFLECTION & NEXT STEPS
This project helped me strengthen my design strategy skills by working within institutional goals while still advocating for student needs. It also pushed me to think differently about designing for AI tools, which are still an emerging space. I learned how important it is to make powerful AI capabilities feel understandable, trustworthy, and approachable, especially for users who may not know where to start. Even though the scope was limited by the class timeline, redesigning MiMaizey pushed me to think more critically about how to communicate value, surface capabilities, and create an experience that feels approachable.
Moving Forward…
I want to dive deeper into the AI capabilities behind MiMaizey to design features that feel even more personalized and adaptive. This could include creating task-specific workspaces (like studying, drafting, organizing sources, etc.), enabling more customized student profiles, and exploring ways to integrate additional UM services and campus systems directly into the experience. Expanding these connections would allow MiMaizey to become a more holistic academic assistant, maximizing the capabilities of AI to deliver smarter, more context-aware support across a student’s entire workflow..