Pitt’s AI Cloud Innovation Center is scaling up and branching out

Research projects in academia often take years. But in Pitt's AI Cloud Innovation Center (CIC), solutions happen in weeks. There, a team of students and staff solve thorny problems brought by Pitt faculty and staff, and students receive funding and training to use the latest in Amazon's AI toolkit.

"We really view this as a workforce development program for the students," said Kate Ulreich, Pitt Digital's lead for digital marketing strategy. "This is giving them the tools, and experience with the tools, that they are going to work with. It's so much more than a job."

The center is a collaboration with Amazon Web Services (AWS), which provides free access to tools and funding for students, along with an on-campus AWS solutions architect for assistance. Pitt's CIC - currently one of only three CICs supported by Amazon and the only one on the East Coast - has found unique success thanks to Pitt Digital's AI expertise and strong initial collaborators in the schools of the health sciences and Pitt Athletics.

"We just had the right mix of being able to put all these pieces together," said Joe Yun, the CIC's leader and chief AI architect for Pitt Digital. "We've got a really strong team at Pitt Digital that was able to bring a lot of foundational understanding."

What began as a one-year contract focusing on health sciences and sports analytics has been extended and expanded in scope, incorporating more students working across a wider range of disciplines. The center is even open to taking projects from outside the University, like local government and nonprofits.

Now in its second year, the center is in one way a victim of that success, according to Yun: They can't hold onto students, who quickly get snapped up for competitive new roles and internships thanks to their hands-on experience with the latest cloud-based AI technologies.

"Companies are just hungry for people that have this skillset," Yun said. "These students are hot commodities."

To succeed in the CIC, students need both technical chops and soft skills. Developing bespoke technical solutions for researchers involves lots of back and forth - not to mention presenting their projects - and means students can't slack on their interpersonal skills.

"We're looking for more than just a coder in the corner," said Ulreich. "We look for students who are very well-rounded," prioritizing problem-solving, communications skills and drive.

Take, for instance, computer science and data science junior Gary Farrell (pictured above). A drone aficionado who was building computers in high school, Farrell is no stranger to the technical side. But he's not content to stick to coding - he's taken rhetoric and public speaking classes to earn a business minor, and he didn't balk when the CIC presented him with a project in the unfamiliar-to-him discipline of health care research.

Hospital systems encode information in different ways, depending on the record-keeping vendor they use. That's a pain for researchers like School of Medicine Associate Professor Christopher Horvat, who, when studying this data, must spend hours painstakingly matching records together before they can start their research in earnest.

Horvat brought his request to the CIC in fall 2025, and right after his initial training, Farrell took it on as his first project. The first thing he received was a spreadsheet with 40 million rows of codes.

"It was definitely eye-opening to try and learn all these new terms and concepts," Farrell said.

But he met regularly with Horvat and went to CIC staff for support and in just eight weeks was able to write a program that performed that matching process automatically with high accuracy. The prototype Farrell created, as with all the projects in the CIC, was then made available to the public - free and open source for others to use and modify. It could be a substantial time-saver for medical researchers, and Horvat is working on deploying it himself.

"It was remarkable how quickly he was able to learn these all of these terms and come up with some ideas for how to how to attack this," said Ulreich. "It was very cool to watch."

After his experience in the CIC, Farrell took a summer internship at PPG working on cloud platform engineering and, having returned in the fall, is now on his fourth project for the CIC. He's covered an array of topics, from x-ray alignment for orthopedic surgery to learning-time analysis in Canvas: A day at the center is as likely to find Farrell reading a medical journal as coding in a terminal (or, more likely these days, directing AI tools to do the low-level coding while he plans out the bigger picture).

All the while, he's building his skills with AWS tools and in working with clients with their own needs and expertise.

"There's always something new to learn, either with the domain or with how we want to solve it," Farrell said. "It's a fantastic networking experience, being able to meet all these different people and learn about all these different professions and companies."

That's representative of the experience students earn in the CIC - a combination of the latest technical tools and a supportive system for using them - something they would be hard-pressed to find in any typical classroom.

"There's no training facility that you could pay for that would give a student this kind of experience that makes them this competitive in the market," said Yun.

Photography by Aimee Obidzinski