Data Utility in the Intelligence Community

The panel entitled, “Elastic Data: Improving Data Utility for People and Machines” consisted of

• Mike Barretta: Manager of IC Solution Architecture, Elastic (Moderator)
• Doug Cossa: Chief Information Officer, Defense Intelligence Agency (DIA)
• La'Naia Jones: Deputy Chief Information Officer, National Security Agency (NSA)
• Saurin Shah: Vice President, Booz Allen Hamilton (BAH)

Cossa explained that for DIA data utility is defined in terms of its value for intelligence function and domain environment. When evaluating data and its utility, he looks for data access and discoverability, data pedigree, data curation, and data leveragability, meaning low barriers to entry for bringing it into the environment.

According to Jones at NSA, they look at flexibility, agility, and the ability to create data that is defined in such a way that it can be used across the various landscapes.  The ability to use the data across the landscape increases its value.  Jones said that as data continues to grow, for her agency it’s not just about the expanse of the data, but how they can make smarter use of it.

According to Shah, as BAH assists federal clients with their IT needs, when it comes to data utility, the question is not about the return on investment of a single piece of data, but the integration of several data sources to drive the answer to a question. The data needs to be discoverable, accessible, processable, and the results need to be explainable.

The IC launched a cloud initiative in 2012 called the Intelligence Community Information Technology Enterprise (ICITE) concept. ICITE aimed to provide intelligence professionals, wherever they are, with the ability to efficiently and effectively discover, access, and exploit data.

During the panel discussion, the moderator asked Jones and Cossa to discuss what principles guided decision-making for ICITE modernization. Jones said that NSA really embraced ICITE around 2018. They are working toward common tools and applications.  Big areas of emphasis at the time were ICAM (Identity, Credentialing, and Access Management) and cloud.  Through this effort, they’ve been able to bring in COTS products to make some processes easier.

Cossa explained that ICITE has been an evolution within the IC. The “cloud-first” concept has led DIA to move most data to the cloud. Some of their data will never be in the cloud due to its significance, such as nuclear control. Lessons learned include efficiencies, security and economics. They had to determine what was necessary on-premise and what was not. They also learned that it makes sense to keep the data close to where it is being used. There’s a time and cost component to transition data outside from on-prem. On the analytics side, machines give the impression they are neutral, according to Cossa, but that’s not necessarily the case.  Machines automate patterns that can be biased. DIA and the IC in general, are in the process of determining what functions could be replaced by automation and what critical thinking needs to be retained. They need to use machines to support more mundane functions.

According to Shah at BAH, one area that has been taking hold is DataOps which involves operationalizing data workflows.  Shah said there are three requirements for DataOps:

• Culture change
• DataOps approach - which means applying an agile methodology to data, using evolutionary data architectures, and building data pipelines.
• Interoperability – such as building APIs

Cossa said that one of the biggest gaps for maximizing the use of data at DIA is on the infrastructure side. To date, most of the focus has been on the needs of the user in the form of capabilities and analytics needed. But they’ve neglected to focus on the foundational technology to support those capabilities. JWICS (Joint Worldwide Intelligence Communication System) is focused on starting at the foundation.  Metrics are another area they need to bolster. DIA needs to determine how to measure performance, the effectiveness of investments, and how to best measure qualitative factors in addition to quantitative.

To close the session, Barretta asked panelists to describe an ideal high utility data environment. Cossa said that a few years ago, he would have said shared strategies, but due to the work that Jones completed when she was at ODNI in the way of developing technology roadmaps and enterprise licensing agreements, this is no longer the case. He said that data use and the technology surrounding it will be a continuous evolution. Ease of use and eliminating barriers for using technology and data will also designate a high utility environment. 

Jones went on to say that at NSA, their most prized asset is the people. Expertise, skills, adaptability to accomplish the mission are paramount. She described needing equal partnerships from vendors.  Not only does the IC need vendors to work with them, but they need vendors to also work together with each other.