Editor’s note: This article is based on insights from a podcast series. The views expressed in the podcast reflect the speakers’ perspectives and do not necessarily represent those of this publication. Readers are encouraged to explore the full podcast for additional context.
On the “CAIO Connect” podcast, host Sanjay Puri sat down with Bharat Prabhakaran, VP & Chief Digital Officer at the University of Cincinnati, to explore a surprising truth: one of the most ambitious AI transformations in America isn’t happening in Silicon Valley, it’s unfolding on a college campus.
In this conversation, Prabhakaran shared how enterprise-grade AI thinking is being adapted to a mission-driven, governance-heavy academic environment and why student success, not profit, is the north star. Prabhakaran’s journey is defined by calculated risks and continuous learning.
After arriving in the U.S. as a graduate student, he chose corporate technology over academia, navigating the dot-com era and spending 18 years at Oracle in multiple roles. Instead of job-hopping, he rotated internally, building depth and breadth.
But mid-career, purpose began to matter more than scale. That pivot from Oracle to finance and eventually to higher education was about impact. The lesson? Career acceleration isn’t about speed. It’s about flexibility, networks, and the courage to realign when your values shift.
During the discussion, Prabhakaran candidly described the culture shock of moving from enterprise tech to academia.
In corporate settings, KPIs revolve around profit and market share. At the University of Cincinnati, the metrics are retention, student performance, and graduation rates. Governance is consensus-driven, not top-down. Relationships and trust matter more than hierarchy.
The technical skills transfer. The mindset must evolve.
READ: Tukatech CEO Ram Sareen on building a business like a family (April 1, 2026)
A major highlight of the conversation was UC’s creation of its own Azure OpenAI instance; Bearcat GPT, “Bearcat GPT is just our instance of Azure OpenAI. So, we’re partnered with Microsoft.”
Why build internally? Data privacy, intellectual property protection, and research sensitivity.
Rather than a campus-wide rollout, UC piloted the system with roughly 1,000 users. A call for use cases generated nearly 100 ideas, narrowed to 14 through a structured rubric. Four governance committees: Teaching & Learning, Responsible AI, Policy & Guidelines, and Operations & Enablement, oversee adoption.
The message is clear: AI in institutions requires governance, enablement, and grassroots buy-in not just deployment.
One of the most compelling initiatives discussed on the podcast was UC’s AI tutoring tools, including Bearcat Study Pal.
Instead of simply giving answers, the AI uses a Socratic design asking guiding questions to strengthen understanding. The long-term vision? Personalization tied to Canvas LMS data, adapting to each student’s learning history.
Prabhakaran acknowledged a hard truth: AI tutoring research is mixed. Dependency is a real risk. Intentional design, not just technical capability, determines success.
Prabhakaran framed AI ROI in two buckets: operational efficiency and academic outcomes, “I know we’re not for profit, but we’re still a business, if we don’t make money, we can’t support our students… and, we have some metrics for those areas where the ROI is not really measured in dollars and cents.”
READ: ‘Responsible AI isn’t optional’: Artur Turemka on scaling trust in a fragmented world (April 3, 2026)
Academic ROI means better retention and fewer withdrawals. Operational ROI includes automation wins, such as invoice processing improvements familiar enterprise use cases adapted to campus realities.
The key insight he shared was speak your stakeholder’s language. Faculty care about student outcomes. Finance leaders care about efficiency.
Agentic AI is coming to higher education but data readiness is the bottleneck.
At the University of Cincinnati, initiatives like Bearcat Insights (a lakehouse architecture combining AWS, Snowflake, and Informatica), “an enterprise data strategy… the building of our Lakehouse, a modern data platform.” The aim is to unify siloed data. Only then can intelligent agents meaningfully support enrollment workflows, advising, and back-office operations.
Prabhakaran closed his conversation with a timeless reminder that technology is the easiest part and mentioned, “…learn and be flexible… You are as good as your network.”
Changing mindsets is harder.
The future of AI in higher education won’t be determined by platforms alone. It will be shaped by leaders who connect stakeholders, communicate clearly, and align innovation with institutional purpose.
In an era of AI acceleration and enrollment uncertainty, universities that thrive will be those that remember: people first, process second, and technology third.