Mahwah High School’s schedule is one of the more complex structures in secondary education. The school runs what’s called a rotating drop schedule. There is a different order of periods every day, and two periods have one “off” day each four-day cycle. Adding further complexity: each science course is paired with a consecutive lab period once per cycle, and PE courses are paired with science on the days when the lab doesn’t run.

Within this structure, the school manages a set of additional constraints familiar to many districts:

• By contract, special education co-teachers require shared collaborative prep periods
• The in-school therapeutic program requires both a morning and an afternoon period
• Students in offsite CTE and work-based learning programs must be scheduled before any other courses are placed
• Remediation courses can only run in the morning
• Students can request overrides into AP courses after the schedule is already built, meaning section counts can shift after staffing decisions are made

Managing all of these constraints manually created a lack of visibility and led to departments working in silos, unable to understand the impact of their decisions on the schedule as a whole. It was standard practice for department supervisors to come to multi-day scheduling meetings with hand-built spreadsheets, and present those numbers to the Board of Education for staffing approval, typically before the complete student course request picture was clear.

Without a way to model what a section decision meant downstream, the consequences only showed up later: a wave of AP override requests would leave some sections overloaded while others sat half-empty, and the ripple effect on teacher loads were invisible until it was too late to act. The result was a dramatic imbalance across the entire schedule. Some PE sections were running with 150 students while others had only 35.

When assistant principal Craig Jandoli and the Mahwah team started working with Timely, they came in with two goals. First, they wanted the process to be easier and to save time. Second, and most important, they wanted to build a schedule genuinely centered on students: their course requests, their priorities, their access to the classes they needed. 

The scheduling team spent the fall translating Mahwah’s constraints and priorities into scheduling parameters that Timely’s optimization engine would honor when auto-generating staff and student schedules. When it came time to start building scenarios, they were finally able to crack the science and PE pairing problem using an iterative approach: first running the optimizer without PE to lock every other placement, then running it again to fill PE into the remaining slots.

Beyond solving the structural constraints, Timely gave the Mahwah team something they hadn’t had before: the ability to show department heads exactly how their decisions impacted the entire schedule.

Mr. Jandoli described Timely’s assigned staff view as one of the features he valued most, particularly the ability to ask “what if” questions and know whether a scenario was even feasible before committing to it. Using that view, he could demonstrate that choosing two sections instead of three meant one teacher would have 180 students while another had 110. Decisions that had been guided by years of tradition were now grounded in data.

The student-centered shift also introduced a new practice in how counselors worked with students on elective selection. Rather than treating all elective requests as interchangeable, counselors had students rank their elective preferences in priority order and fed that information directly into the optimizer so Timely could prioritize placements accordingly. 

With each optimizer run, Mr. Jandoli and his team could experiment with different scenarios, locking certain placements, adjusting section counts, and running the optimizer again to test the impact. “It allowed me to really narrow in on what combination of constraints would make the best schedule. I’ve never experienced that before in scheduling,” he said.

Mahwah finished the season with over 20 percent more students having full schedules than in years past. Simultaneously they were able to satisfy 97 percent of all core course requests. The PE imbalance that had defined previous years was resolved, and class sizes came out balanced across all grade levels and subject areas.

Mr. Jandoli was candid that there’s still room to improve, and equally candid about where the work lies. “A schedule is only as great as the data that we provide to it,” he said. “I’m confident that we can get a higher percentage of students fully scheduled before hand-scheduling. That’s not a Timely issue, that’s our issue. And now we know exactly what to fix.” Working with Timely has made visible what years of manual scheduling had obscured: where their own internal processes could be tightened to give the optimizer even better inputs to work with.

Beyond the numbers, the team felt most proud of what their new scheduling process allowed them to create. Before working with Timely, Mahwah’s schedule was “more of a teacher-driven, what-human-resources-do-I-have type of schedule,” Janodli said. “Whereas now this really was driven by course requests. The analytic piece of it, being able to see what kids actually wanted and what kids prioritized, was really enlightening to us.”

The shift registered with counselors too. When Craig previewed the schedule with them, “they were blown away by the ease in which the optimizer placed the sections where they belong when we allowed requests to drive the schedule.” The balanced class sizes reinforced that impression, and when it came time to hand-schedule students with unresolved conflicts, counselors were happy to complete what had previously been a dreaded and time-consuming task.

Working with Timely also changed how the Mahwah team thinks about the scheduling timeline itself. In years past, the work had to begin in February simply to leave enough time for the manual process. In practice that meant building a schedule with incomplete data and revising it multiple times as staff retired, IEP needs shifted, and waiver decisions came in late. 

For next year, the plan is to wait and use the extra time to perfect their scheduling data inputs. “Knowing that the optimizer can get me a schedule, getting student requests met and balancing the bazillion constraints that we put on the schedule, in 90 minutes or less, that’s where we’re going to start seeing the time-saving benefits,” Mr. Jandoli said. Cleaner data going in means fewer rounds of rework coming out, and a schedule that better reflects the reality of the school by the time it’s built.