IE 11 Not Supported

For optimal browsing, we recommend Chrome, Firefox or Safari browsers.

It’s Still Early, but Process Mining Is a Hit in Oklahoma

This growing form of getting more from data is helping officials analyze spending that took place without top-level oversight. The tool could find its way into the hands of other public agencies for a variety of tasks.

Data Shutterstock
The state of Oklahoma has embraced a relatively new form of data analysis called process mining, and its experiences could guide how other governments find ways to squeeze more value from procurement and otherwise become more efficient.

The technique is called process mining.

The technology — in this case, provided by Celonis, the first time a U.S. state has used its process mining tools — uses specialized algorithms that analyze data so users can better understand agency operations, workflows and spending.

As Celonis likes to put it when marketing the software, “think of an MRI that shows how your processes actually run — not how you think they run.” The general idea is to improve government procedures and delivery of services, saving money when possible.

Oklahoma’s Office of Management and Enterprise Services (OMES) earlier this year deployed the Celonis Execution Management System. It came after a 2022 review of state agency purchases, with lawmakers telling OMES to increase control over spending, and do so quickly, Blaine Bridges, the state’s assessment and standards manager, told Government Technology via email.

That happened after the state identified more than $3 billion worth of procurement in fiscal year 2022 that apparently went through without approval from OMES Central Purchasing. Some agencies can make their own purchases thanks to exemptions, the number of which reportedly have increased to 87 from seven.

Even so, state officials reportedly were spooked by the spending review.

Without the new platform, that state-mandated call for more insight into that spending would have been done manually via an ERP system. Bridges called that a “painstaking endeavor requiring cross-reference on multiple platforms and a significant amount of manual work hours.”

As Bridges described it, the first step in the OMES process mining effort was defining the state’s purchasing standards within the platform. Then the technology can map “the actual trajectory of purchases to identify anomalies and flag anything that falls outside those standards.”

The result: more insight into deviations from the typical or ideal spending process, which in turn can lead to fixes and improvements.

Using process mining, OMES has reviewed more than 16,000 purchase orders with a total value of $2.26 billion.

“Previously, processing such large amounts of financial and procurement data within the time frame needed would have required us to contract with a third party of approximately 45 people,” Bridges wrote. “Now, our small internal team can perform these duties utilizing the tools at hand in a matter of weeks.”

The work also has shed light on exactly how agencies define those exemptions for their purchases compared with how state law is written, Bridges said. As well, process mining widens the view when it comes to agencies and data sets, providing more opportunities for change.

“The resources are at our fingertips in real time and in an easily accessible format within the platform,” Bridges wrote.

Process mining can take on added importance with the recent rise in inflation and for states facing such financial roadblocks as deficits, rising health-care costs and swelling pensions, Colin Wardlaw, vice president of public sector for Celonis, told Government Technology.

As well, COVID-19 showed how easily some government systems can be overrun with demand and data — another argument in favor of process mining, he said.

The technique is designed so that users can “look forward” and deal with real and potential problems that way. Wardlaw added that process mining is relatively easy to scale. And while doing it well typically requires the participation of government employees who know data and business analysis, as the experience in Oklahoma is showing, process mining teams don’t have to necessarily become big.

Process mining seems destined for a bright future in Oklahoma, at least according to Joe McIntosh, the state’s chief information officer.

"This technology solution not only drives efficiency but also transparency, accountability and compliance,” he told Government Technology via email. “We're better equipped to make real-time, data-driven decisions that benefit taxpayers, agencies and suppliers."
Thad Rueter writes about the business of government technology. He covered local and state governments for newspapers in the Chicago area and Florida, as well as e-commerce, digital payments and related topics for various publications. He lives in Wisconsin.