The next phase for ASU will involve embracing AI beyond research, bringing the innovative potential of generative AI to teaching and learning using Google Cloud. 

ASU Prep is the university’s free public charter school network that reaches 10,000 K-12 students across Arizona through its network of in-person, hybrid, and digital public charter schools, as well as 30,000 students internationally in collaboration with other schools. Collaboration with Google Public Sector supports the development of new educational models that help students succeed. One project involves developing immersive learning simulations for everything from anatomy to government. Built on Unreal Engine and running on Google Cloud, the simulations give students a new way to look at learning. 

Their most recent ASU Prep project, Archie, leverages Gemini and other generative AI models to develop a support chatbot for math problems. As middle school and high school students attempt their math work, Archie is by their side to identify errors, provide hints and guidance, and help explain problems in real time. Working with the hint and error feedback functions in the tool increased the number of times that students got the right answer on their first attempt by 6%—a big step in building confidence and setting up students for success. 

“The Google Public Sector team really embraced the idea of using AI to help students through different styles of learning,” says Amy McGrath, Vice President for Educational Outreach and Student Services, and Managing Director of ASU Preparatory Academy at ASU. “They put everything towards creating models that put students at the center.” 

The ASU EdPlus team encourages diverse student populations to reach their educational dreams through ASU Online. Online student populations tend to be highly diverse and often include learners who are returning to the classroom or starting at an older age. These learners are more likely to be juggling jobs and families, making higher education a big decision.The EdPlus team wanted to better support students with effective outreach, and that began with a comprehensive review of every interaction. 

Working with Google Public Sector, the EdPlus team built an AI powered platform that ingests audio recordings from enrollment and student success centers (which support all student needs, from initial curiosity with ASUO through graduation) and converts them into text. The transcripts are processed to extract student insights securely and within compliance. 

The EdPlus team used web behavior data to build a propensity model that predicts the likelihood of someone enrolling in ASU Online. The new enrollment propensity model is four times more accurate than naive model guessing enrollment status at random, allowing EdPlus to concentrate resources on supporting and encouraging high-propensity prospects. Marketing campaigns using this model are also more effective, increasing registrations by 52% compared to traditional lead scores. The next iteration of this model, which has already begun development, will incorporate all the qualitative conversation data from the enrollment and success centers ASU and Google Public Sector are partnering to accelerate research, engage students, and support education for all. Together, they are co-developing solutions to address pressing and evolving challenges in education and research. ASU team members provide valuable feedback to Google Public Sector, helping to shape the future of education and research technology.