U-M GPT’s Maizey Tool Guides Students to Success
One year ago, the University of Michigan became the first university worldwide to launch an exclusive AI platform for its community. Today, students and faculty at Michigan Ross are using the new generative AI tools to enhance learning inside and outside of the classroom.
Technology and operations professors Jun Li and Andrew Wu faced a common problem that plagued professors in many different subjects: overcrowded office hours, during which students sought guidance from professors and teaching assistants. With eight sections totaling 600 students and only so many hours in a week, Li and Wu explored the new generative AI tools at U-M to create a solution for their problem.
Through their collaboration, Li and Wu trained U-M’s Maizey to create an AI tutor for the Operations Management 313 class. Acting like a 24/7 teaching assistant, students can ask the AI tutor questions about the subject matter and receive an answer that uses information from the course content. It even directs students to areas within the course materials where details can be found.
“One of the key things that I hoped to achieve was that I didn’t want the tool directly spoon-feeding the final answer to the students. That would defeat the purpose of the learning process,” Li said. “But I do tell the AI tutor to help promote independent thinking from the students. So, the tutor’s goal is not to provide a final answer, but actually help with the thought process.”
The Maizey AI tutor has increased efficiency for both instructors and students. According to Wu, the tutor saved instructors five-12 hours of answering questions each week. For students, the tool provided an ability to ask questions beyond the classroom or a professor’s office hours. This was a key benefit for students trying to balance their schedules.
Lydia Willett, a BBA ‘25 student in the Operations Management 313 class, credits the Maizey AI tool with helping her review course materials in a more efficient manner.
“With the AI tool, I could ask questions and get immediate answers at whichever time I was studying,” Willet said. “This helped me go through the material in a more efficient way; I wasn’t stopping-starting-stopping when I had a question and, as before, having to email the professor or move on while feeling unsteady in some answer that I had arrived at on my own.”
This increased efficiency was evident in Willett and her classmates' performance. According to Li, in-class quiz scores significantly increased after implementing the Maizey AI tutor.
Due to its recent emergence, some students have been hesitant to utilize generative AI tools. However, the more they use them, the more they recognize their benefits.
“When I was told that we would be able to access an AI tutor that was trained and tested on the course’s materials, I was, at first, nervous to use it. To me, the course was already difficult, and I was scared I’d get really complex, technical solutions returned to me by AI,” Willet said. “I thought when talking to someone face to face, like in office hours, at least I’d be able to express the extent that I already understood the lectures.”
To address some of these concerns, Li and Wu modified the tool so that it does not generate instant answers but instead refers students to the course materials where they can find the answers on their own.
“I was studying for the final exam when I finally used the platform, and, almost immediately, I wished I had done so earlier,” Willet said. “I wasn’t getting any crazy, complex responses; they were straightforward and complemented the explanations I struggled with in my own notes. If I was confused about something, I could ask the tutor to explain an answer in another way, or by using simpler language, or then check if I was correct in my understanding of the AI tutor’s response.”
As generative AI evolves, Li and Wu continue to search for ways to integrate it into classrooms to enhance learning and productivity. They have submitted a grant proposal to expand access to the AI tutor throughout the more than 30 statistics courses offered at U-M.
“We will continue to expand aggressively in integrating genAI into our education and research initiatives,” Wu said. “Our students and alumni can look forward to a variety of exciting new courses, activities, and opportunities to get involved in the coming semester.”
More information about U-M’s Maizey tool.
