Behind every government mainframe computer is an untold story.
Rarely do we hear about the employees who dedicate their careers to writing code, building systems and finding ways to keep the government’s critical, bulk transactions running for decades. These stories don’t get the glory and rarely get attention unless something goes horribly wrong.
But dissecting these stories can unlock truths about the past that help us understand the present. They can also help us press past the hype of modern versus legacy and focus on the real issue at hand: data and how it’s accessed and used in ways that affect everyday people.
Terry Milholland, the first-ever Chief Technology Officer and Chief Information Officer at the Internal Revenue Service, knows this firsthand. He is credited with heralding an “IT renaissance” of people, processes and technology in the IRS.
Speaking at GovLoop’s recent Briefing Center, a two-hour collection of different online trainings, Milholland described the multi-pronged approach that the IRS took to free up data buried in legacy systems.
“It’s the kind of message that all of you can take to heart as you all undoubtedly have legacy systems where data is trapped,” Milholland said. He was joined by Chris Borneman, Chief Technology Officer at Software AG Government Solutions, who explained how the company’s technical solutions have helped the IRS and other government agencies free up data, respond to legislative demands and improve services.
One of the big lessons that Milholland shared was this: If we want to get access to the data we have to start with developing a data model. It should be the sole, authoritative repository of the data. For the IRS, that was critical for having a single source of truth about tax information.
“But the real issue underneath all this is people,” he said. The people who maintained legacy systems have a vested interest. They build their careers around it, and they were dedicated to the mission.
“We had to find a way to enroll that community into the new [way of working and thinking],” he said.
That exercise began with getting people who were closely tied to the legacy systems to help design the data model. Milholland’s goal was to help them see the bigger picture. The work they were doing was to support taxpayers across the country — a mission that was bigger than any one person.
The focus was on creating a world-class organization, and that meant acting like any well-managed private sector organization despite the funding and political challenges that come with being a government agency.
“It enabled people to say, ‘I am not just working on ancient stuff, I am helping the IRS free up data,” Milholland said. The agency provided training for employees and standardized around a data model and programming language. While the IRS worked to replace legacy systems over time, application programming interfaces (APIs) became the intermediary for connecting and retrieving data across 700-plus IRS systems. APIs are software that dictate how systems interact with each other, and they allow the public to interact with some of the government’s backend systems.
As the number of APIs crept up, the IRS had to also create standards around APIs, ultimately requiring that they use an integration platform by Software AG called webMethods.
Oftentimes in government, tech investments are born out of legislative necessities, but those requirements come with little to no budget. Based on legislation such as the Affordable Care Act, for example, the IRS turned from being a tax collector to an information provider, Borneman said.
Many other agencies and commercial businesses rely on tax services for record validation and to determine who qualifies for which payments, he said. Data is getting more complex at a time when there is greater demand for data and transparency.
The COVID-19 relief funds are a prime example. Different states have requirements for wage garnishments, which must be taken into account before funding is disbursed. Today, the IRS is better positioned to respond to these types of requirements.
Borneman offered this advice for agencies seeking to free up data that’s trapped in legacy systems:
- Understand what you’re trying to accomplish and the definition of success.
- Show the value. Start with small use cases that address your biggest pain points.
- Share those successes and always keep scalability in mind. How will you scale whatever success you’ve demonstrated?
- Approach your work with the right architecture, or blueprint.
This online training was brought to you by:
