Building the Data Governance Strategies of the Future

The Status of Data Governance in Government

he Government Accountability Office (GAO) receives varied data and data structures for oversight purposes, and must have the capacity to analyze and support these data sets to develop credible oversight products.

Taka Ariga, chief data scientist and director of the Innovation Lab, Science, Technology Assessment, and Analytics for GAO, said the agency is addressing its data strategy in three components: data literacy, data science and data governance.

“For us, data governance starts with a notion of cataloging all of the data assets coming into the agencies,” Ariga said. Then, the proper rights and access privileges associated with each data set are applied. “Even though we have the statutory access to that information, we want to make sure that we are absolutely safeguarding the security and privacy entrusted to GAO.”

The Virginia Data Commission leveraged the state’s data governance strategy to drive real-time data analytics and decision-making during COVID-19. The commonwealth had to identify how health systems were operating and determine whether any specific health regions were overwhelmed with cases. It leveraged its partnerships with private health care member associations to access those data assets and turn them into actionable intelligence. The data showed that the commonwealth didn’t need to stand up any additional health care facilities and should reallocate that effort to testing.

“That was an immediate thing that we could do based on just the analytics that we were seeing, the results and the intelligence we were getting,” said the commonwealth’s CDO Carlos Rivero. By having a strong data governance foundation, the result of connecting the right data, insights and algorithms allows agencies to optimize processes, increase efficiencies and drive innovation.

Behind the Chief Data Officers Council

The Federal CDO Council has already set its goals for 2021, yet challenges persist around resource constraints and risk aversion, especially as it pertains to data sharing and ownership. The beauty of the CDO Council, however, is its ability to open communication across government so members can share best practices, solutions and challenges.

“Folks have a lot of competing priorities, and the more we lend each other a hand, the better we all are,” said Greg Fortelny, CDO for the Department of Education. “The CDO Council provided that opportunity that I didn’t have before to develop a common understanding of challenges.”

Agencies can also share the solutions they’re using to improve data management. Once the right foundational data governance and security are in place, agencies can focus on the platforms that best transform that data into insights.

According to Nick Psaki, principal engineer of the Americas – Federal for Pure Storage, the technologies on the horizon that are best suited to help CDOs are high performance and highly-responsive data service infrastructures and next-generation data architectures that provide things like fast block storage and cloud-native software-defined data services.

“This allows invisible and infinite scalable data services that can be tailored as-a-service and acquired as a subscription service,” Psaki said. “This is a boon to data officers.” Rather than specifying an architecture, CDOs can define an effect. This can transform how data platforms are designed and how data is consumed, expanding agency capabilities.

Data Sharing to Make a Difference

In efforts to demonstrate the efficacy of using data to solve complex problems, the commonwealth worked with the Virginia Department of Criminal Justice Services to find a solution to the opioid epidemic.

“We realized that data sharing has been a significant problem,” Rivero said, and addressing the opioid epidemic couldn’t happen without integrating a wide variety of services. So, the commonwealth created a coalition of statewide partners and stakeholders to develop the data sharing and intelligence gathering capabilities needed. They worked together to develop a scalable infrastructure in the cloud so the solution could be expanded to other communities.

The Department of Education was also recently tasked with a massive data collection project when Congress appropriated billions of dollars through the Education Stabilization Fund. The bulk of those funds go to education agencies, governors’ offices and institutes for higher education, and the Department of Education was in charge of ensuring the right money got to the right people at the right time.

With those funds came reporting requirements, so the department had to start up, in record time, a public transparency portal to display data in a way that made sense to customers. This portal was also a data collection mechanism with a built-in data quality edit while making information available to the public in an automated fashion.

“This was really only achievable with the right kind of technology. But perhaps more importantly is the skilled staff that could take that technology and the data to produce a product that really stays true to the complexities of these programs,” Fortelny said.

Preparing for Data Management Strategies of the Future

As organizations adopt artificial intelligence, machine learning and advanced technologies to help make more sense of their data, they’ll also have to scale data governance strategies. “ML models deliver results at scale, but can only be precise when the data they’re feeding on is quality,” said Aileen Black, senior vice president of federal at Collibra.

Data quality is a top barrier to adopting technologies like AI and ML. “You really need to have that in your data governance and your data intelligence platform to really accelerate your pipeline of data,” Black added.

When good data and good models are unified, AI and ML tools can deliver trusted results. And trusted analytics with AI drive more effective decisions, higher productivity and cost savings.

To truly solve government challenges with data, Psaki said agencies must embrace the new, communicate the vision and outcome clearly at all levels, and choose great technologies and technology partners.

“There’s in truth no new problems in the realm of IT, intelligence and information management,” Psaki said. It’s worth asking, “how has this been solved anywhere else in industry or anywhere else in the world, and can we adapt it to our own organizational needs?”

Download a PDF Copy of This Mission Brief