The holiday season is here and with it, the many associated symbols and traditions. One of them, in countless living rooms, has a new twist at Drexel College of Computing & Informatics (CCI) graduate student Anthony Parone’s home this year.
Anthony Parone
He took lessons from his recent class, INFO 629: Applied AI taught by CCI Professor of Computer and Information Science Rosina Weber, PhD, home for a decidedly seasonal application:
“I immediately thought about my holiday train setup,” Parrone said, and how he might use applied AI and large language models (LLMs) to “see if I could let family and guests control it just by chatting.”
Reflecting on the class structure, Weber explained its focus on practical AI applications.
“This class teaches an AI application’s structure the students use to learn about applications developed by others, learn about the AI methods and AI tasks that are part of AI applications and develop their own,” said Weber.
For Parone, it took “about a week of working with Claude (Anthropic AI assistant and LLM) and tweaking the details to get everything running smoothly,” testing it with his teenage son.
The results have been a hit with family and friends. His guests have used the chatbot about a dozen times to control the train’s sounds, lights and movement.
“People love the novelty, especially kids. I expect usage will increase as we have more guests this season,” he said.
Weber emphasized that this hands-on approach is central to the course’s mission.
Rosina Weber, PhD
“This course is primarily about teaching students how to think in terms of applying AI to problems, which is what facilitates and motivates developments like this one,” said Weber. “Tony’s application shows the value of the learning process from this class. The tools are freely available to everyone, but without learning to approach AI applications in a structured way, it is difficult to gain value from them.”
In his CCI classes, Parone has enjoyed “how the professors break down complex AI concepts into understandable pieces, and the hands-on assignments let me immediately apply what I learn,” he said.
He has plans to connect an LLM to his home automation system or start a computer vision project to identify visiting birds at the family’s feeder next.
Beyond individual projects, Weber highlighted how industry trends are shaping the tools students use.
“As part of the AI applications’ structure, we emphasize the benefit an application has to its end-users. Through assignments, we encourage students to evaluate applications even if all resources are not available,” said Weber, continuing, “Tony included the main items from class assignments, including evaluating it with his son. He also describes the open-source large language models (Ollama) tools he used. There is a growing number of tools and frameworks becoming available in support of developing applications particularly in conjunction with large language models like Claude and ChatGPT.”
“The benefit is that these tools are built with commercial-grade software engineering that decrease the risk of bugs at implementation time. As big tech companies compete by offering their tools, we can expect to see more and more of plug-and-play AI applications accessible to those without highly developed coding skills,” she said.
These developments underscore why Parone chose to deepen his expertise through Drexel’s graduate program, enrolling through Comcast’s sponsored program with CCI (he is a Comcast Executive Director of Global IT, Strategic Planning) and returning to the school where he earned his undergraduate degree.
“Coming back with work experience gives me a different perspective on the material,” Parone said.
With the fundamental outlook in place, Weber described what’s next for Parone’s cohort.
“Now that students can think critically about what matters when building AI applications, they are prepared to create solutions with practical value,” said Weber.
For Parone, who expects to complete the degree in 2028, these lessons have the clear purpose of preparing him for leadership in a rapidly evolving tech landscape.
“It'll help me lead technical teams at Comcast more effectively as they adopt these technologies. Understanding the fundamentals lets me make better decisions and communicate with stakeholders,” he said.