CASE STUDY
Kent State University Rebuilds Its Banner Schedule in 26 Days with JAMS
Kent State had 26 days between the contract award and the day its old scheduler expired. The team rebuilt all 182 job chains and went live with no contingency days.
26 days
From kickoff to production cutover
182
Job chains rebuilt by hand
0 days
Contingency in the schedule
Education
North America>United States
Windows, Linux
Banner, AppEnhancer, FormFusion, ePrint
Financial Transactions, File Transfers
About the Company
Kent State University is a public institution with eight campuses across Northeast Ohio and more than 33,000 students. Its student and financial systems run on Ellucian Banner. Behind Banner sits a scheduler that runs the ordinary machinery of a university: financial aid file downloads from the federal servers, population selections, loads, transfers, and the reporting that follows them. Roughly 25,000 job executions run in a normal week.
The Challenges
Kent State had run the same commercial scheduler for more than a decade. By the university's account, the cost rose roughly 150 percent over three years, which pushed the contract past the threshold at which university procurement policy requires a competitive bid. In spring 2026, Kent State issued a request for proposal for an enterprise workload automation and application scheduling platform. The requirements were straightforward. Run the Banner workload, integrate with AppEnhancer, handle Windows and Linux, and do it without a long professional services engagement. The incumbent did not submit a proposal. After more than a decade in the environment, with every job definition on campus written to its format, it let the bid pass. Kent State notified JAMS of the award on May 21, 2026, and issued the purchase order in June. The existing contract expired on June 28. Everything had to be running on the new platform by then, and the team who would do the work had never opened JAMS.
The Solution
The migration started with an export. Kent State pulled an extract of every definition out of the old system, and JAMS ran it through its migration tools, which read those definitions and generated the equivalent JAMS jobs. That gave the team a starting point instead of a blank screen. It did not give them a finished schedule. Kent State uses a one up parameter that was required by additional definitions inside the sequence, in some cases more than once in a single job, so all 182 chains were rebuilt rather than lifted across. Kickoff was June 2. About a week went to servers, databases, agents, and firewall rules. Two days went to training. That left roughly two weeks of building, carried by four full time staff with a Banner developer helping on errors, all of it alongside their normal responsibilities and with a JAMS consultant supporting them. As with any migration, it was not frictionless. Jobs did not always save on the first attempt during the build. A Windows path was being interpreted as a Linux path, which produced errors that made no sense until someone traced them. Global variables took several rounds of back and forth in mid June. None of the roadblocks were showstoppers.
The Benefits
Kent State cut over to JAMS Scheduler on June 28, the day the old contract expired. The Banner workload has run through summer term and into the fall semester. There have been no failed or missed runs attributable to JAMS. What has come up has been additional configuration on individual jobs. The clearest change is in promotion. Building a new chain in the old system took anywhere from fifteen minutes to over an hour in development, and then the same work again to stand it up in production. In JAMS, moving a chain to production takes minutes. Kent State has also given functional users read only access, so the team no longer sends detailed screenshots to get approvals. The team has since configured additional schedulers, requested more training, and started writing its own internal documentation. On August 3, Kent State joined a working session with another Banner school partway through the same migration, and showed them how it passes aid year and term code into population selection jobs at runtime. Six weeks after go live, a team that had never opened JAMS in May was teaching its peers how to run it.
It does everything the old scheduler did, and it gives us more control over how things run. It is far easier to move things from development to production, which is not a small thing for a team our size.
Amanda Kelley-Hurm
Senior Manager, Application Administration Support, Kent State University
What does Kent State get with JAMS?
182 job chains rebuilt and in production in 26 days
Zero contingency days against a hard contract expiry
No failed or missed runs attributable to JAMS since cutover
Chain promotion to production down from up to an hour to minutes
Read only access for functional users, removing screenshot approvals
Roughly 25,000 job executions a week across eight campuses
