By Ingrid Nuttall and Jessica Liebowitz

Every registrar knows the feeling. Bottlenecks in the course registration system mean that students are not able to get the courses they need when they need them. At Northeastern, a global institution spanning 13 campuses in the United States and Canada, students are encouraged to explore across disciplines and campuses. That mobility makes the bottleneck problem an especially high-stakes priority for the university. The prevailing assumption has been that it comes down to seat counts: too many students chasing too few spots in popular courses, while other courses sat half-empty. That assumption held until we got the chance to dig into the data. Doing so required integrating information from three separate university systems: enrollment data in one, course catalog, program structure, and prerequisite information in another, and degree audit data in a third. The resulting insights reframed the problem entirely.

The bottlenecks were not where we thought they were. The most congested points in the curriculum were not the most popular courses. They were courses appearing in multiple degree programs that students across a wide range of majors were required to pass through in order to graduate. These were courses sitting at the densest nodes of dependency, meaning those with the largest number of prerequisites feeding into them, that in turn unlocked the largest number of courses downstream. Students were not stuck because too many of them wanted the same seat. They were stuck because the curriculum was funneling them through mandatory traffic jams.

Questioning the prevailing assumption was made possible by an unconventional collaboration with students that bought us the time, attention, and data scientific know-how necessary for exploration. That initiative is the Student-Administrator Partnership at CPS, which pairs technically proficient students with university administrators to co-develop solutions to operational problems. These students are selected, with the help of participating faculty, for several key qualities: technical excellence in data science and AI, and their ability to teach themselves new skills on the fly as they accept the responsibility for navigating ambiguous problems effectively. Students arrive without solutions and with eagerness to learn about the problem. Administrators hold the operational wisdom to know the details of what needs fixing and the curiosity to experiment with technological innovations that could do the fixing. The initiative’s faculty mentors oversee the rigorous commitment to intellectual creativity and accountability expected of the student partners.

Questioning the prevailing assumption was made possible by an unconventional collaboration with students that bought us the time, attention, and data scientific know-how necessary for exploration. That initiative is the Student-Administrator Partnership at CPS, which pairs technically proficient students with university administrators to co-develop solutions to operational problems.

With the results of the data explorations in hand, our student partners built a “bottleneck dashboard” to help college leadership, faculty planning curriculum, and registrar staff visualize the curriculum network and identify the highest-dependency courses. They then created a “similarity score” showing potential alternatives to each bottleneck course, by applying AI techniques to reviewing every course description in the catalog. Both of these solutions are early-stage, proof of concept tools. Formal confirmation of the structural pathways that characterize each degree program will require feedback from academic administrators across the university. Actual curriculum changes authorizing alternatives to required courses will require faculty deliberation and approval. The point is that thanks to customized data science/AI applications, we now have concrete new ways of thinking about the bottleneck problem that were not visible before.

In the longer run, we expect the infrastructure of these new tools to migrate to a central technology office to help ensure sustainability: in this way, the registrar retains responsibility for interpreting the output and acting on it, while the analytical tools themselves become part of the institution’s operational culture.

Dramatic improvements in university operations do not necessarily require million-dollar technology investments up-front. They can begin with a partnership between an administrative team that is expert in a specific operational problem and one or two students with the technical skills to dig into it, supported by faculty mentorship of the student partners’ intellectual creativity. In our experience with the Registrar’s office, the tools emerging from this kind of collaboration will be prototypes, not ready-made solutions. But that may be its most unique contribution: creating the conditions for administrators to investigate a problem they already know exists but have not had the time or technical resources to get deep enough to see new ways around it.

---

Ingrid Nuttal is Deputy Registrar at Northeastern University. Dr Jessica Liebowitz is Faculty Director of the Student Administrator Partnership at Northeastern’s College of Professional Studies.