Best listening experience is on Chrome, Firefox or Safari. Subscribe to Federal Drive’s daily audio interviews on Apple Podcasts or PodcastOne.
2020 is shaping up to be a big year for the intelligence community and its efforts to take cloud modernization to the next level.
The goal is to develop a clearer narrative and “north star” for the IC’s cloud modernization strategy, John Sherman, IC chief information officer, said Wednesday at AFCEA NOVA’s IC IT Day in Herndon, Virginia.
The IC published an official cloud strategy last year, which consolidated nearly 300 priorities from 17 different agencies into one document.
Now, two major procurements will build on that IC cloud strategy, Sherman said.
The first is a follow-on to the commercial cloud services (C2S), an Amazon Web Services effort that runs on all three security fabrics: top-secret, secret and unclassified. The current model allows the IC to pay only for the cloud services it needs, with Amazon shouldering most of the costs of capital upkeep, patching and deployment of new capabilities.
The next iteration, called “commercial cloud enterprise” or C2E, will build on the successes of C2S,” Sherman said.
C2E will be a multi-cloud, multi-vendor environment, again covering three security levels, he said.
The CIA, which is the lead on this particular effort for the IC, released a draft request for proposal earlier this month for the commercial cloud enterprise, with the final RFP due next month, Sherman said. The IC anticipates it will make initial awards in 2021 but may issue an offer as early as this fall.
“While we’re going to have to operate in this new environment, just as we learned how to operate with a single commercial cloud at all three security fabrics, we’re very optimistic about our forthcoming opportunities to leverage, for example, best-athlete capabilities, in a multi-cloud environment and in an ecosystem that can encourage multi-vendor competition,” he said. “This is the right thing for us in the intelligence community at the right time.”
The next iteration of the GovCloud will known as the “hybrid compute initiative” (HCI), which Sherman described as a “high-performance analytic environment.”
“HCI will be tailored to address the massive processing and analytical requirements that NSA has with software that must run at very honed specifications on bare metal hardware in cloud compute that must be bought in bulk, given the data quantities and processing demands at NSA,” he said. “HCI supports a mission at NSA that is always on, all the time and running at very high capacity.”
The vendor for HCI will be responsible for providing the majority of the infrastructure and hardware, which can’t be too far from NSA’s facilities, Sherman said.
“With this requirement, of course, comes opportunity,” he said. “This new model for managed-service hosting we envision will enable NSA to vacate quite a bit of its legacy data centers and similar IT hosting facilities on their premises.”
NSA plans to issue RFPs for HCI hardware-as-a-service and bulk cloud compute next month, with the goal of an award later this year, Sherman said.
DLA to restore order to its ‘data house’
Cloud is also a priority for the Defense Logistics Agency, as DLA plans to lift and shift some of its 150 existing systems to the cloud this year, George Duchak, the agency’s CIO, said.
The plan, he said, is to run two geographically separate instances of those systems. Once they’re in the cloud, DLA will begin to rationalize those systems.
“We’re on a journey to unify approximately 150 standalone, siloed systems into a common platform for our customers … for our suppliers, all 12,000 of them, and for our users, and to provide them with a common user experience and a personalized user interface,” he said. “But before all this, we have to get our data house in order. Our data today is scattered and stranded in those individual 150 systems, preventing us from fully realizing and exploiting the power of artificial intelligence.”
After all, DLA has seen some progress with the emerging technology pilots its launched and the commercial AI technology it’s purchased over the last year.
DLA, for example, began to embrace robotics process automation near the end of 2018, Duchak said. It deployed 20 bots that year and another 53 bots in 2019.
Together, those bots are projected to contribute nearly 130,000 man hours a year back to the agency, Duchak said. DLA bots are speeding up the employee onboarding process, handling some Freedom of Information Act requests and providing some financial management support.
The General Services Administration has taken notice of DLA’s initial success. GSA’s RPA playbook, which the agency released last month, uses a four-level model to assess organizations and their maturity with the emerging technology.
Duchak said DLA is the only government organization that sits at level four, which GSA has described as a high-performing RPA program.
All of DLA’s functional bots to date are considered “attended” bots, meaning a human being must monitor and review the automated work.
This year, DLA will focus on deploying “unattended bots,” Duchak said.
“These bots can operate on a schedule, use their own certificates to access websites and systems,” he said. “This means that once they’re programmed, they become digital workers that independently contribute to mission.”