When the Justice Department named a new chief data officer, it didn’t name a position, but a person.
Joe Klimavicz is that person wearing both the chief data officer and chief information officer hats for the Justice Department. He said it makes sense for him to do both jobs, at least for now.
“Serving in these roles is an opportunity to take a holistic and efficient and effective approach to managing data and implementing new requirements. I get to look at information and data from many different legal aspects, through all these different acts and authorities, and address privacy, security, interoperability and data management together,” Klimavicz said on Ask the CIO. “Meeting the data requirements from the new legislation will help us improve our overall IT posture within the department. I can rely on Federal IT Acquisition Reform Act (FITARA) and CIO authorities. They are all mutually beneficial. I think my CIO authorities will help accelerate our data management within the department.”
Agencies had until July 13 to name a chief data officer under the Evidence-Based Policymaking law. When Justice named Klimavicz, some experts expressed concern over dual-hatting the CIO, given how important both positions are and without a singular focus, how much of an impact can a dual-hatted person have in either role.
Klimavicz said in the future Justice may separate the CIO and CDO roles, but for now bringing them under one person makes the most sense.
“Where we are with our focus on the basics of information and data management, and these topics are in my core focus areas as CIO. I think there is a lot of beneficial synergy and I can leverage different authorities from both the information and data side to make significant progress much faster in aggregate,” he said. “There is a lot of similarities [between data and information] in terms of how we look at these. In every department, it’s going to be a little different in terms of how they look at it. Some agencies are very focused on statistical reporting, data collection and so on, and that may drive them in a different direction. In our areas right now, there is a lot of overlap.”
As CDO, Klimavicz said his priorities build off of the Justice Department’s first data strategy, which he released earlier this year. Along with the federal data strategy, he said Justice has a multi-year roadmap to improve how it manages, shares and secures its data.
“We need to build that organizational capacity to build that data culture and enhance the skill sets of our workforce. Data scientists are really hard to find and hire, but it’s a critical aspect that we really need to focus on in our recruiting,” he said. “It all starts with having a data management plan. We have hundreds of different systems in the department and making sure we have plans to effectively manage information and data from a lifecycle perspective. We have to bolster our data standards, access policies and goals, and making sure that’s a shared understanding. We are trying to develop a more holistic or umbrella data architecture that includes shared standards and taxonomy. And application programming interfaces will be critical so maybe build in an API library to enable that seamless information sharing.”
For Justice the culture change is more than standards and processes, Klimavicz said he’s maturing a data community of interest to share best practices, focus on workforce training and developing a more rigorous governance model. One of his goals is to finalize the CDO board chartered and set up this year.
Several DoJ components already have CDOs too, including the FBI, so Klimavicz wants to expand on the work of these data leaders. He also wants to include more than just CDOs in the governance efforts, but bring in other disciplines and expertise, for instance from the Office of Justice Programs, which is built around data.
“The good thing is I think we have a very mature governance model within the department on the CIO side of the house so I think we will try to leverage what we’ve done there and really provide a strong focus on information and data management. We plan to leverage some of the structures we have in place for the CDO activities,” Klimavicz said. “I think coordinating and collaborating with the components will be absolutely crucial to developing these capabilities. One thing I’ve learned is just telling people what to do is probably not the most effective way to get things done. It’s much better to work collaboratively upfront and make sure what you are doing adds value.”
For many, that value must be around using data to make better decisions. Klimavicz said Justice has to do a better job of operationalizing its data by understanding what it has and making sure it’s properly cataloged.
“We need to enhance our analytical capabilities to effectively use this information. That means identifying and linking related data, using artificial intelligence and machine learning, creating visualization and dynamic dashboards, adding geospatial fields because I believe almost all information has a geospatial component to it and employing data science to really bring all that stuff together,” he said. “We are maturing our approach by implementing an enterprise data analytics blanket purchase agreement so we can effectively get to more analytical tools and support the organization in how we analyze large amounts of data.”
Among the benefits of this new contract will be helping Justice bring disparate and unstructured data together. Justice currently has 36 different cloud environments so at least connecting the data repositories is key to making AI and machine learning effective.
At the same time, Klimavicz said the technology is only as effective as the governance and workforce that underlie it.
“Our ability to leverage these types of advanced analytics requires us to have a robust data architecture across the department. Think about it as a platform or a common way of sharing existing and future data sets,” he said. “We also need to lower the barrier of entry of bringing some of these analytical capabilities in. I would prefer to bring in new technologies to work against the data rather than a lot more staff. It’s easier and in the long run it’s probably less expensive.”