MITS Students Harness AI’s Potential to Advance Research and Education
By Lindsay Marcellus
From creating a framework for meaningful feedback in the classroom to revolutionizing prospects for increasing energy production securely, artificial intelligence (AI) continues to pose opportunities for transforming how we teach, learn and respond to pressing challenges.
At Carnegie Mellon University, students in the multidisciplinary Master of Information Technology Strategy (MITS) program are tackling the challenge of deploying AI tools to advance research and education across disciplines in the university setting. Through a series of capstone projects, MITS students are helping to enhance the ability of organizations to make effective decisions in an increasingly complex and rapidly changing world.
Partnering with Organizations to Solve Strategic Problems
Serving as the culmination of the MITS program, which is administered by the Carnegie Mellon Institute for Strategy and Technology (CMIST), the capstone project enables students to apply their classroom knowledge by partnering with a company, nonprofit or educational organization to help solve a strategic problem. For nearly a year, student teams have worked with Dietrich Computing and Operations on an ongoing series of projects that focus on how humans and humans working with AI can make decisions in dynamic scenarios. The results of this AI-facilitated practice in decision-making are poised to have an impact on academic disciplines ranging from cybersecurity to economics, as well as on the fields of management and leadership.
Last fall, in a project that received the Top Capstone Award at CMIST’s 2025 diploma ceremony, Jacob Lapin, Phillip “PJ” Roach Jr. and Jonah Nascimento focused on developing a more accurate, data-driven model for assessing national power. Applying frameworks they learned during their MITS coursework, the students modeled industrial pivotability using public datasets and then extended that into a strategic wargaming use case. To address limitations in widely used statistical measures such as the Composite Index of National Capacity, the team built a neural network-based model that analyzes industry-specific data — such as workforce size, worker activities, products, assets, and capital-labor ratios — which assess the feasibility of one industry pivoting to support another.
Left to right: Phillip “PJ” Roach, Jr., Jonah Nascimento, and Jacob Lapin present their capstone project in fall 2024
“It was rewarding to work on a problem that combined the technical skills with analysis that could have real-world applications,” said Nascimento, who added that the exposure to national security concepts through his coursework helped him to understand national power from a broader perspective.
Simulating Complex Scenarios
While the results of the “Rethinking National Power” project can serve as a standalone index for comparing key facets of national power, it also can be leveraged for advanced adjudication in complex simulation scenarios through the Operational Gaming Environment (OGE). Part of a comprehensive toolkit to strategically and responsibly integrate AI into university settings, the OGE is one application within the Dietrich, Analysis, Research, Education (DARE) Platform.
Under the guidance of Vincent Sha, associate dean of information technology and operations for Dietrich College, two other MITS student capstone teams have collaborated with DARE to explore the use of large language models (LLMs) to transform educational and simulation-based learning.
Last fall, Yifan Luo, Yintong Zheng and Jiaji Li focused on two applications: Socratic Books and OGE. Applying their software development and AI-related knowledge, the students developed the user interface and created live web app demos and gameplay walkthroughs. The project offered an opportunity to apply material learned in the MITS curriculum to projects designed to facilitate interactive learning in the classroom.
Yifan Luo and Yintong Zhang work on a project in 2024
“The OGE Platform is a specialized layer of tools that accelerates analysis, facilitates strategy development, enhances decision-making abilities and improves negotiation skills, all while reducing the organizational cost of creating and managing the event,” said Rafael López, CMIST deputy director for security policy studies. López anticipates that as the tools mature, they will be helpful not only to students and instructors, but also for organizations seeking to better understand their industry environments and planners and leaders who are looking to better appreciate risks. “These tools have a direct application to real problems,” he said.
For their final MITS project, Daniel Hayase and Alex Wan spent the summer refining and testing the OGE ecosystem, contributing improvements to game logic, UI/UX design and AI advisor behavior. Their “Quick Start Guide for Instructors” aims to help faculty members streamline lesson preparation. “If we can empower instructors to tailor scenarios to their curricula and give students the agency to explore complex problems safely, I think we’ll spark deeper engagement, sharper critical-thinking skills and a lasting appetite for collaborative problem-solving,” said Hayase.
For both Hayase and Wan, the opportunity to improve the platform through finding new bugs and improving the user experience and user interface was a highlight of the capstone project.
“I’ve really enjoyed testing the AI advisor bots and finding creative ways to make them hallucinate. It feels like solving a puzzle with constantly shifting rules,” said Wan.
Left to right: Alex Wan and Daniel Hayase present their capstone project in July 2025
Their work supports CMIST’s dedication to the wise development, use, and governance of new and emerging technologies.
“Experiments by student teams helped us to catalog and understand the implications of hallucinations and data accuracy when blending real-world and fictitious data. This understanding is critical in developing reliable scenarios for simulations,” said Sha.
Applying Capstone Project Experiences To Careers
In addition to enjoying the collaborative work that integrates and applies what they have learned in their CMU coursework, MITS alumni report that the capstone project experience positively impacts their career after graduation.
“This capstone project sharpened my ability to distill complex technical data into actionable insights for senior decision-makers,” said Lapin, who now serves as a Cyber Warfare Officer in the U.S. Army. “It also improved my collaboration and project management skills,” he added.
Luo highlighted the value of practicing communicating technically complex work clearly. “It is crucial to learn to explain why and how tasks were completed and the value of your work,” she said.
While these projects all have immediate applications for responsibly integrating AI within the classroom, the experience has also helped students gain an understanding of the impact of emerging technologies on international and national security and strategic decision-making.
“It has given me a better appreciation for some of the challenges developers face daily and has highlighted the importance of aligning technical decisions with broader strategic goals,” said Hayase.
The Master of Information Technology Strategy is a cooperative endeavor administered by the Carnegie Mellon Institute for Strategy & Technology and in partnership with the College of Engineering and the School of Computer Science. To learn more about the program, or if your organization is interested in sponsoring a capstone project, please contact Mark Gardner, CMIST’s graduate program manager.