Michelia Mason didn’t start out as a data expert. In fact, when she first opened Power BI, the data visualization tool felt more intimidating than intuitive. “Data analysis was not my background,” she admits. “Usage reports were about it.” But as the librarian at Southwest Research Institute (SwRI), a nonprofit based in San Antonio that supports everything from space science to automotive fuel research across 10 technical divisions, Michelia knew her library needed a better way to communicate its impact. So, she got curious—and started experimenting with AI.

Michelia’s story is a reassuring one for academic and public librarians who might feel unsure about how to begin using data or integrating AI tools into their workflow. With no formal training in Power BI or AI, she simply started asking questions. “If I didn’t know how to do something, I’d ask ChatGPT,” she said. “It would give me a step-by-step DAX expression or Power Query action I could plug into Power BI. If I got an error, I’d feed that error back into ChatGPT and it would help me troubleshoot.”

That back-and-forth with AI became a practical, hands-on way to learn—what she calls “a journey that’s still ongoing.” Over time, Michelia has used ChatGPT and Power BI to build dashboards that help answer key questions, like:

• Are we maintaining equitable distribution of interlibrary loan (ILL) requests across responsive libraries?
  
• Do ILL request trends reveal consistent content gaps in specific research areas?
  
• What is the prevalence of Open Access scholarship in peer-reviewed publications by research area?

Data as Advocacy

One of Michelia’s biggest takeaways is that data isn’t just about measurement—it’s about storytelling.

“We want to move beyond cost-per-use,” she said. “We want to understand our impact. What are researchers publishing? Which journals are they citing? And how much of that content is the result of our subscriptions?”

In one analysis, she found that 88% of peer-reviewed citations in published research were directly supported by library-provided content. That’s a powerful stat when advocating for funding or demonstrating return on investment.

She’s also starting to collect more qualitative data and documenting time spent with researchers, publications supported, and department-specific engagement.

Her team now pulls in data from Tipasa (for ILL analytics), Scopus (for bibliometric analysis), and COUNTER reports (for e-resource usage) to create dashboards embedded in SharePoint. While nothing is fully automated yet, Michelia has created and used code, in the form of Python scripts, again with AI’s help, to streamline how monthly data is uploaded and organized.

You don’t need to be an expert in Power BI or AI to get started. You just need a question. “I had a vision in my mind,” Michelia said, “and I used ChatGPT as an assistant to help me bring it to life.”

SwRI is a special library, but she has advice for librarians in any setting. “Own your data. Start with the story you want to tell.” Whether it’s demonstrating the value of your services to funders or identifying trends in patron behavior, tools like Power BI and ChatGPT can help you speak in the language of stakeholders—data and dollar signs.

And if it still feels daunting? Michelia suggests you “start small, keep learning, and don’t be afraid to try.” After all, she’s still learning too and has a goal to go back to school for data science with hopes of sharing what she’s learned with other librarians who want to do more with their data.

“You don’t have to know it all,” she says. “You just have to start.”