IT Officials Are Finding Efficiencies Across Cloud, Data Networks

With cloud tools and data analytics taking hold across the federal government, both agencies and industry are turning to new methods to optimize operations.

At GovCIO Media & Research’s Federal Cloud & Data Forum last week in D.C., agency and industry leaders highlighted the technology they are using to gain data insights.

CBP Moves to the Cloud

To streamline its operations, ranging from processing 1 million inbound travelers and 250,000 vehicles daily, Customs and Border Protection completed its cloud migration last month, shifting more than 60 applications and databases while continuing to maintain operations.

Adina Pantella, executive director of CBP’s Passenger System Program Directorate, said the migration has already shown efficiency results for the agency.

“In our legacy data centers, handling big surges in travel or sudden operational spikes amid hardware constraints and complex provisioning. In the cloud, we can scale dynamically and instantaneously to meet those mission demands,” she said.

The migration also allows CBP to now operate with little to no downtime, providing the agency the ability to update and upgrade without applications going offline.

Pantella said the shift has also allowed CBP to integrate new artificial intelligence tools the agency had not before been able to access.

“We’re leveraging AI on two fronts: internally through code-assist processes that accelerate development velocity and accuracy, and also operationally by integrating AI-powered tools that assist in identifying emerging threats and streamlining the processing of legitimate drivers,” she said.

Faster Patching to Keep Up with Software Vulnerabilities

A June directive from the Cybersecurity and Infrastructure Security Agency (CISA) has significantly trimmed the time federal civilian agencies have to patch highest-risk software vulnerabilities to three days.

Katie Chrisman, federal solutions engineering team lead at NinjaOne, said the directive BOD 26-04, which adopts a more risk-based approach to vulnerabilities, also gives researchers less time to test the patches before they’re pushed out.

“Here’s the problem: the way most of us have patched for 20 years is stage that patch in a test lab, run some tests against it, against representative systems, ensure nothing breaks, then releases to the fleet,” she said. “That process takes days, sometimes weeks, and it exists for good reason.”

Chrisman noted that a pair of updates pushed out in 2024 inadvertently led to system crashes millions of Windows devices.

She said to allow developers to adapt and safely test patches quickly, NinjaOne has developed models using AI tools to help attain that speed by using early adopters to set the baseline across millions of endpoints and continuously evaluating the patches’ risk scores.

“Three days is not a target to survive; it’s a new baseline for everyone, not just by federal agencies,” Chrisman said. “The organizations that get there won’t be the ones that patch the fastest. They’ll be the ones who patch fast and also trust the underlying data.”

CMS Tames the Multi-Cloud Environment

For Wade Zarriello, director of the Centers for Medicare & Medicaid Services’ Infrastructure and User Services Group, cloud capabilities are no longer the aspiration of agencies, but the tools for officials to optimally deploy.

At CMS, that meant developing a cloud ecosystem where multiple networks could operate in collaboration.

“Instead of asking how do we migrate an application, we’re asking how do we help customers make better decisions before they migrate? How do we govern that consistently? How do we provide operational financial intelligence? And how do we make financial management part of engineering rather than an afterthought,” Zarriello said. “Those questions became the blueprint for CMS Cloud Fusion ecosystem.”

The CMS Cloud Fusion ecosystem pairs networks from several cloud service providers, as well as other on-prem networks, fused with tools that help integrate workloads across multiple cloud environments.

Zarriello said the integrated tools have helped CMS better manage its workloads and governance across cloud networks and better manage its spending through unified business intelligence.

“What excites me the most isn’t simply that we built these capabilities, it’s how they’re beginning to work together,” he said. “From isolated tools to connected experiences, from fragmented processes to integrated workflows, from reactive support to proactive engagement, customers increasingly experience one coordinated ecosystem with multiple cloud providers and functionality, not siloed environment ecosystems.”

NIH’s Integrated Data Platform Reaches Across Datasets

At the National Institutes of Health, officials are working across cloud platforms to try and achieve the holy grail of federal data, dismantling siloes to gain more insights.

The agency is attacking the problem with something called its Integrated Data Platform (IDP), a centralized tool that integrate datasets from across NIH.

“At its core, what IDP is meant to do and is trying to do something very simple that you guys have heard over and over today, which is break down data silos,” said NIH Lead Management Analyst Steevenson Nelson. “IDP is meant to serve as a data gravitational point for NIH data, clinical data, genomics data, research data, you name it.”

Nelson said NIH has been also working with CMS to partner on some data sharing that can glean new insights through the IDP.

“So if one thinks about NIH as 27 different institutes and centers that will have all of these different data types across business areas, with all of that sitting on one place, which you guys have heard today … AI really is heavily dependent on good, clean data that is centralized, and that’s what we’re actually aiming to do with IDP. And I think we’re going to get there.”

DIA Sifts Through Open Intelligence

To capitalize on the insights that publicly available information may provide, the Defense Intelligence Agency is working to collect and make sense of the proliferation of what’s known as open source intelligence (OSINT).

Eric Miller, DIA’s senior advisor for open source intelligence, said collectors within DIA are constantly scanning public channels like social media to find quality OSINT informing military intelligence abroad.

“The best lesson I learned from day one is that when something happens and there’s a crisis, senior leaders ask one thing, ‘What does OSINT have to say about it?’ Which was a fascinating kind of humbling for me to see, in part because we can produce information in hours and minutes. We can turn things around incredibly quickly,” he said.

Given the scale of OSINT being produced daily, Miller said DIA is looking to automate its collection and make the data more accessible.

“We have data sets, for instance, that have tens of millions of data points in them. No person can do anything with them, right? We need systems that can help us process that, and that’s one thing we’re working on as well,” he said.