Close

New Research from CDW Explores AI and Cybersecurity

Learn how AI is helping IT teams manage risk and improve resilience.

Nov 09 2026
Artificial Intelligence

How Higher Ed Institutions Are Using Vibe Coding To Solve Problems

Two institutions share how artificial intelligence–driven software development can streamline processes.

Many colleges and universities rely on custom software to address unique campus workflows or data privacy needs. Because of limited IT resources, relatively few have historically opted to build those applications on their own.

Now some are predicting that’s about to change as “vibe coding” within higher ed moves from niche to mainstream. The software development method entails describing an app in plain language and letting an AI assistant generate the code. The approach isn’t foolproof, those with experience say. But with practice and attention to data governance and security, it can be an effective way to solve problems on a budget.

“It’s so powerful and so achievable, it’s amazing to me that more people aren’t using vibe coding already,” says Brian Basgen, vice president of digital and physical infrastructure and CIO at Emerson College. The technology “isn’t perfect,” he adds, and it does come with certain risks, “but if you’re thoughtful and careful, there’s no reason not to try it, if only to see what it can do.”

Click the banner below to find data governance solutions for AI applications.

 

Practice and Patience Lead to Success at Emerson College

At Emerson, Basgen’s team has used vibe coding in Claude to develop nearly two dozen apps in-house. 

“When the institution comes to us with a challenge,” he says, “the first thing I think about now is whether we can use AI to solve it.” The answer depends on the confidentiality of the data they’d need to make any solution functional, and it hinges on the hypothetical software’s overall importance to the college. “Anything highly complex, we’re not going to do it; if it’s available commercially, we’ll go with that instead.”

One recent problem that did meet the team’s requirements was Emerson’s need for a more effective way to manage hundreds of test cases during a student information system implementation. Normally, Basgen explains, with many people testing, they’d use spreadsheets and pivot tables to track the results. 

DISCOVER: Customized software solutions help universities meet institutional goals.

“We were like, ‘Wouldn’t it be nice if we just made a simple web app so that everybody could access everything easily?’” Leveraging what he describes as their vibe-coding infrastructure — a standardized development environment created in Claude — they quickly built a solution that fit the bill. “We did it in almost no time at all, and it’s substantially better in terms of insights and tracking than the traditional implementation process.”

Other recent vibe-coded successes for the team include a security review app that helps staff understand Workday access configurations and troubleshoot missing permissions, and a department budget tracking app that integrates with financial data through Snowflake.

When Basgen and his colleagues first experimented with vibe coding in the fall of 2025, they found the technology difficult to work with and prone to making mistakes. Through trial and error, they eventually got up to speed, and now they have a repeatable process spanning everything from change control to code audits and security risk assessments.

“The No. 1 thing was getting started,” Basgen says. “After that, it was just persistence, patience and vision.”

READ MORE: A cyber resilience program should examine possible risk.

Vibe Coding at the University of Montana

Another IT leader who’s bullish about vibe coding is the University of Montana’s Zach Rossmiller, associate vice president and CIO. For the past several months, Rossmiller says, UM IT has hosted weekly “Vibe Fridays” to give campus employees dedicated time to explore how vibe coding might benefit them. They discuss their biggest challenges, test potential solutions, and for staff who don’t know where to start, his team provides guidance and encourages experimentation. “We’re trying to show them the possibilities,” he explains. “Like, here’s this AI tool, see what you can do.”

His team recently helped the UM Summer office use vibe coding to build a dashboard to analyze enrollment and estimate which courses might generate revenue versus costing the campus money. And for their own use, IT created a new systems status page after finding that a commercial product designed for that purpose included licensing restrictions that severely limited its utility. “After a couple hours, we had a prototype up and running that did what we needed it to do,” Rossmiller says.

Like Basgen, Rossmiller stresses the importance of treating vibe coded projects the same as any software requisition. Before they moved their status page app into production, for example, they put it through a battery of tests, including rigorous scanning and vulnerability assessments. And for any potential project that relies on access to the university’s data, “the first thing we’re going to do is talk to the data steward,” he says.

UP NEXT: Building an AI program requires a solid data strategy.

The way he sees it, Rossmiller adds, AI can produce a functioning app, but it can’t guarantee that software is going to be secure, accessible and maintainable. 

“We’re very cautious with what we do and how we do it,” he says. “For us, that’s the only way to put vibe coding to work.”

Luis Alvarez/Getty Images