.svg)
While the enterprise world scrambles to deploy AI, some leaders are choosing a different path—one defined by deliberate patience over speed. In higher education, where the core mission is to foster student success, the calculus for adopting new technology is different. Instead of rushing into the AI add-ons pitched by software providers daily, some institutions are fostering a culture of experimentation, recognizing that the biggest risk isn't being left behind, but making hasty investments with limited resources. At Hamilton College, AI integration and adoption isn’t being dictated from the top down or forcing the various departments into a single enterprise solution. Instead, the college is fostering a culture of experimentation, letting innovation emerge from students, faculty, and staff in tandem with their "What If?" campaign.
This philosophy is championed by Michael Sprague, the Director of Enterprise Information Systems at Hamilton College. With over 20 years of experience, Sprague oversees the college's complex ecosystem of enterprise software and integration tools. Unlike many institutions that feel pressure to implement vendor-driven solutions, Hamilton’s IT team is intentionally keeping its options open.
A warning on vendor hype: "Avoid just going with whatever your current vendor tells you to go with. Nine times out of ten, they’re bolting AI on because it’s the buzzword of the moment, but it’s not always the best choice," Sprague said. "In many ERP products, AI is not their area of expertise, so I try to avoid getting locked into something to make sure we explore newer options with a more open mind."
Big SIS: I have heard of other schools that automatically go with their big SIS (student information system) vendor's solution, and it's not always the best choice for the outcomes departments or institutions are trying to achieve. I'm not saying they're all bad; I'm just saying IT leaders need to explore and do proper research before committing."
For Sprague, the demand for AI is unlike any previous technology adoption cycle. In the past, implementing enterprise software required a significant internal sales effort to generate buy-in. With AI, the energy is inverted. From his vantage point, The college’s AI momentum is largely grassroots, with ideas flowing in from departments across campus.
Sprague points to initiatives like the college’s "What-If Initiative" as central to this energy as AI appears in numerous ideas and proposals, "It's a changing environment for us and a good time to be working at the college through this period of change."
An experimental mandate: "Maybe eventually, as we continue to develop the strategy, we'll move toward having one solution that's our default and then some others for various use," Sprague said. "But it's so early right now that we really don't want to lock down, because six months from now you could have a totally different solution with the way things are moving. So we just continue to prototype, try things, and learn."
An organic shift: "Enterprise software is pretty boring. You have to generate the buy-in, you have to sell it and convince people why they have to have it," Sprague explained. "But with AI, it's coming naturally from every angle. We're not out there driving it; we're more on the service side of it."
This approach is made possible by a key structural advantage inherent to higher education. Free from the relentless pressure of quarterly ROI that drives public companies, institutions like Hamilton have the space to test, learn, and iterate without demanding immediate financial returns. This allows for a more fluid and flexible definition of "production."
The freedom to prototype: "Maybe it's an advantage of being in higher ed, but we're experimenting more," Sprague noted. "Most of these tools you can switch in and out; they're not like the traditional enterprise software where you've got to do a two-year implementation."
Risk-managed experimentation: "We wouldn't go into our office of fundraising, for example, and force an AI solution and say, 'Here's how it all works,' and just go all in 100% on it," he said. "We'd still do a prototype with a small group of donors and experiment to see what we can do with it. We would not implement something that's going to force that across everything unless we know that it's a successful endeavor."
This culture of widespread experimentation introduces an obvious tension: how do you empower everyone to innovate without creating a chaotic environment of Shadow IT, Shadow AI, and other security risks? For Sprague, the answer lies in the robust governance foundation his team has already built, with the goal being to support people exploring AI while ensuring they do so safely.
Freedom within a framework: "We have a really strong governance backdrop already from a lot of the integration work and system implementations we've done," he explained. "To us, this is more about helping people and reacting to the needs and the ideas they have that might make improvements anywhere at the college. We're able to use our pre-existing governance backdrop to be more nimble with some of these ideas and help people experiment and prototype various things."
The leadership paradox: When asked if it was frightening to know that students, faculty, staff, and even administrators were exploring tools before IT even found out, Sprague's answer was candid. "It does, but that's the wrong philosophy to have toward this," he admitted. "I know that it's part of my job to make sure everything's compliant and follows all the regulations and policies. I can still do that, but I have to keep an open mind to what all these different solutions are."
The ultimate validation of this strategy isn't found in a strategic plan, but in the tangible outcomes it produces. The hands-off, supportive approach is fostering a campus-wide explosion in AI literacy, empowering non-technical users to become builders.
Low-code/No-code: “We’re seeing staff, faculty, and students using AI to find the solutions themselves. You don’t have to be a coder, you don’t have to have an IT background,” Sprague says. He highlighted a recent student project: “I saw a student group, none of whom were coders or computer science majors, create a recycling bin for clothing and build an app that goes with it entirely through AI. They just built the thing in two days.” While the prototypes are usually rough and elementary, the capability is impressive. “It was rough around the edges, but just to have the capability to do things like that—I think it’s pretty amazing right now.” He adds, “Most of the time if students are going to try and build out an MVP or project prototype, they would hire a developer and refine it. But just to have the capability to do things like LCNC and vibe coding."
IT's new role: What's most fascinating to Sprague is the new workload he and the rest of the IT team are starting to encounter: students asking for help, and not for traditional help desk items. Sprague shared, "Now we have folks coming from all corners of campus to IT to ask if we can look at these projects or refine it for them. It's a great way to have conversations with our campus community about security, governance, and other items that usually aren't top of mind in an approachable and relevant setting."
By combining bottom-up experimentation with a strong governance backdrop, Hamilton College is learning, iterating, and positioning itself to adopt AI strategically while avoiding premature lock-in. As Sprague puts it, “We don’t want to be a roadblock for AI innovation. We want to learn from what everyone’s doing across enterprise and education to equip our campus to be agile enough to know what's the best path forward.”
