New Era of AI Assistants at the University
When the Compass Assistant and Resource Advisor (CARA) recently went live, it marked more than the launch of another chatbot. It represented something bigger: a growing, campus-wide movement to bring practical, secure, and deeply integrated AI tools into everyday work across the University of Arizona.
At the center of that momentum is Jay Timsina, Enterprise IT Architect for Academic, Student, and Business Operations, who has emerged as a driving force behind the ArizonaAI platform. But as Jay is quick to point out, this work is anything but a solo effort. With strong support from Mark Felix, Executive Director of Enterprise Business Applications, and the UITS leadership team, the initiative has continued to grow and evolve. “We are offering an AI-powered chatbot service built on the ArizonaAI platform – a custom solution developed by University Information Technology Services (UITS).”
CARA, built for the College of Agriculture, Life and Environmental Sciences (CALES), is designed to help faculty and staff quickly access guidance on everything from business services and reporting to organizational structure and awards. It’s a practical tool, grounded in real needs, and powered by a platform designed to scale.
Data Scientist Benjamin Markwart, working on data solutions for CALES, saw an opportunity early on to address a long-standing challenge: online information existed, but it wasn’t easily accessible. Through close collaboration with Jay Timsina and the ArizonaAI platform, that idea evolved into a fully realized assistant, one that required not just technical integration but careful iteration, testing, and refinement to ensure it delivered meaningful, trustworthy results.
Behind the scenes, collaboration is what makes the ArizonaAI platform work. UITS contributors, including Sameer Khanal, Web Application Developer, Senior; Ron Bradley, Web Application Developer IV; David Pichevin, Software Engineer; Seth Pederson, Senior Analyst, Implementation; and Mario Vasquez, User Experience Designer, helped build the technical foundation, while partners across colleges and units shaped how each assistant serves its audience.
The ArizonaAI platform enables departments, business units, and individuals to create virtually unlimited, highly customized assistants tailored to their goals and data. Timsina notes, “It also ensures strong governance, supports custom evaluation benchmarks for each assistant, and provides robust observability.”
That platform integrates seamlessly with existing university systems and supports a wide range of use cases, from fully public-facing assistants to tightly restricted, domain-specific applications. Markwart explains about his work on CARA, “It wasn’t just building a chatbot. It was working through how it connects to real university systems, how it handles access, and how it performs reliably. That’s where the collaboration with Jay and his team really mattered.”
The AI Assistant interfaces with existing authentication and authorization systems, including WebAuth, Microsoft Entra, LMS authentication, role-based access control, student/instructor enrollment verification and local authentication for global users. Each use case can also maintain its own data pipelines and isolated vector stores.
CARA is just one example. Across campus, a growing ecosystem of AI assistants is taking shape. Each assistant reflects a partnership between UITS technical teams and business owners, who understand the day-to-day challenges and opportunities in their subject areas.
- Tools supporting innovation and commercialization through Tech Launch Arizona
- Assistants improving student services in Housing and Student Unions
- Simulated clinical environment assistant helpsstudents to build real-world skills
- Graduate and Payroll assistants streamlining complex administrative processes
- Business Intelligence and Data Solutions assistant guides users in navigating reporting systems
“Each assistant reflects the people building it,” notes Markwart. “The ArizonaAI platform gives you the foundation, but the real value comes from how you adapt it to your users.” The result is a shared capability that enables the university to move faster, respond more effectively, and deliver better experiences across teaching, research, and business operations.
For Jay, that’s the real story. “It’s about enabling many teams to create solutions that fit their needs while contributing to something larger.” As more units adopt the platform, the vision becomes clearer: a connected network of AI assistants, built by the university for the university, each one strengthening how work gets done and how people connect with the information they need.