For years, colleges and universities have invested heavily in technology with the promise of improving efficiency, insight, and student success. But according to Joe Abraham, co-founder and CEO of Intellicampus, the problem facing higher education today isn’t a lack of technology — it’s too much of it, scattered and disconnected.
“Most institutions are running 80 to 120 different systems on campus,” Abraham said, drawing on more than a decade of consulting work with presidents and senior leadership teams at universities and large community colleges. “All of which have value, but none of which talk to each other.”
That fragmentation, he argues, has quietly become one of the biggest barriers to delivering the seamless student experience institutions say they want.
When strategy meets reality
Abraham traces the problem back to a familiar pattern. University leaders articulate a clear institutional strategy, often with broad agreement across cabinet-level leadership. But when it comes time to implement that vision, the complexity of campus systems gets in the way.
“You come up with a great strategy with the president and the cabinet,” he said. “But when it comes time to deploy it — especially when you’re dealing with the IT shop — you run straight into headwinds.”
Those headwinds come from decades of incremental technology purchases. Large enterprise systems such as student information systems, learning management systems, and customer relationship management tools sit alongside dozens of smaller, department-specific applications. Each one stores data differently, uses different permissions and operates largely in isolation.
The result, Abraham said, is an experience that feels disjointed for both staff and students.
“To help this student, I have to log out of three systems and log into two others,” he said. “My sign-on doesn’t work here, so I have a different login there. Meanwhile, the student is standing in front of me.”
The student experience problem
Students feel that fragmentation most acutely. A new student may need to navigate separate systems for registration, housing, advising, billing, and coursework — each with its own interface and rules.
“Because the data sits in all these fragmented systems, we’re asking the student to go to the fragmented systems,” Abraham said.
Single sign-on solutions were meant to ease that burden, but they often fall short. “I log into this system, but it logs me out of that system,” students told researchers in focus groups Abraham has conducted. “When I log out here, it times me out somewhere else.”
The impact is not evenly distributed. Students who are already comfortable navigating higher education systems adapt more easily. Others do not.
“Our most at-risk students are the ones more likely to say, ‘Forget it. I’ve got to go to work,’” Abraham said. “And then it doesn’t get done.”
AI won’t fix fragmentation — at least not by itself
As institutions rush to adopt AI tools, Abraham warns that AI layered on top of fragmented systems can actually make things worse.
“AI can only solve problems with the data it has access to,” he said. “If each AI is tied to its own system, then just like the systems don’t talk to each other, the AIs don’t talk to each other either.”
There are also serious governance concerns. Once AI has access to data, it doesn’t naturally respect institutional boundaries.
“It’s very hard to govern AI in an environment that’s built on tables and rows,” Abraham said. “There’s no way to keep it from getting to the next byte of data.”
Without clear rules and unified infrastructure, institutions risk exposing sensitive information or creating unintended access to protected records.
The Case for a Unified Data Layer
The solution, Abraham said, lies in unifying data before expanding AI use. Other industries, he notes, are already paving the way.
“Microsoft and others are building what’s called Data Fabric,” he said. “It takes data from lots of different places and starts to connect it outside the systems.”
Rather than replacing existing platforms, data fabric technology pulls information from them and standardizes how it is understood. A student ID may be labeled differently across systems, but the unified layer recognizes them as the same thing.
“Until you bring that data out and unify it, AI can’t make sense of it,” Abraham said.
This approach replaces older concepts such as data lakes and warehouses, which were designed primarily for reporting rather than for real-time orchestration and AI-driven experiences.
Cost isn’t the real barrier
Despite assumptions to the contrary, Abraham says cost is not the biggest obstacle.
“Believe it or not, it’s a fraction of the cost of replacing an SIS,” he said. While major system replacements can run into the tens of millions, data-unification platforms typically cost hundreds of thousands of dollars spread over multiple budget years.
“The bigger issue is the human side of change,” he added.
Faculty worry about the role of AI in teaching and assessment. Staff worry about job security. Long-standing business processes are challenged.
“A lot of it is helping humans identify where AI can free up their time — and where it’s not needed,” Abraham said.
Leadership, governance and patience
One of Abraham’s strongest recommendations is that institutional presidents retain ownership of AI strategy.
“It’s so transformational that unless the president stays the executive sponsor, the silo effect kicks in,” he said. Delegating too narrowly can skew AI efforts toward academics or operations, missing broader opportunities.
He also urges institutions to resist the urge to chase flashy tools before the groundwork is laid.
“Don’t go buying the faucets,” he said. “Put in the plumbing first.”
That means cleaning data, setting governance rules, and building infrastructure before launching pilots. Progress, he emphasizes, should be measured in years, not months.
“Every school should look at this as a two-to-five-year journey,” Abraham said. “By 2030 or 2031, institutions that do this right can be completely transformed.”
For Abraham, the message is clear: the future of AI in higher education won’t be defined by the tools institutions buy, but by whether they solve the fragmentation problem that’s been holding them back all along.