How federal agency leaders are using AWS tools to improve efficiency

Federal agencies are working to meet demand for systems that can keep pace with a tech-savvy public. Agency leaders are embracing the opportunity to reimagine government by modernizing legacy systems and finding more digitally secure ways to serve. Federal leaders presented at the AWS Summit in Washington, DC 2025 to highlight ways they’re using Amazon Web Services (AWS) to improve efficiency.

Departments are using cloud-based tools to increase productivity, reduce resource use, and improve their ability to efficiently serve the public. This article highlights the pioneering leaders forging new paths into cloud-based government.

FCC modernizes 94 percent of systems

The Federal Communications Commission (FCC) was wrestling with outdated legacy infrastructure that had become costly and inefficient to maintain. Frequent hardware challenges, outdated application tech stacks, and increasing cybersecurity compliance challenges were putting pressure on the agency. High operational costs, limited scalability, and a dependency on specialized skillset to manage aging systems were slowing progress. Don Tweedie, FCC’s deputy chief information officer for technology delivery, and his team recognized that a fundamental transformation was needed.

Tweedie led the charge by developing a strategic playbook to modernize the FCC’s network architecture. His vision focused on a cloud-first approach—prioritizing scalability, security, and agility. The agency began treating data as a strategic asset, implementing micro-segmentation and leveraging cloud-native capabilities like autoscaling and serverless computing. The result was a shift toward data-driven decision-making and a more resilient infrastructure.

In just 20 months, the FCC migrated 94 percent of its data center operations to the cloud and decommissioned around 300 underutilized servers—streamlining its tech stack and reducing operational drag.

Tweedie emphasizes the importance of starting with small, manageable pilot programs to build momentum and organizational buy-in. He recommends involving security teams from the outset, reinforcing zero-trust principles, upskilling existing staff to address talent gaps, and establishing clear, actionable policies to guide transformation.

KCNSC closes its collaboration gap

The Kansas City National Security Campus (KCNSC) works on behalf of the National Nuclear Security Administration as part of the Department of Energy to safeguard the nation’s nuclear program. With a distributed workforce of over 7,000 people, the organization was finding it increasingly difficult to build an effective collaborative environment on the unclassified side. The KCNSC team grappled with inefficient data siloing, lengthy discussions about the use of agency resources, and the inability to scale.

To solve this pressing problem, Victor Doane, enterprise architect with the digital transformation program, championed the idea of a Joint Unclassified Cloud Environment (JUCE). He engaged engineering leaders and other site teams to collectively define the organization’s needs and to discuss their different cloud migration journeys to learn from each other’s experience. Doane prioritized getting leadership on board, then facilitated an agency-wide culture shift toward modernization with the help of AWS Professional Services.

KCNSC used the Landing Zone Accelerator from AWS as a base for its unified cloud environment, then iterated according to their needs. The project took off quickly: Doane pitched the concept and successfully brought it to fruition in less than a year. The agency plans to use the enterprise blueprint to support other department sites on their cloud journeys.

NIF pursuing AI to recover lost work hours

Lawrence Livermore National Laboratory’s (LLNL) National Ignition Facility (NIF) houses some of the world’s most energetic lasers, which initiate nuclear fusion reactions like the ones that energize the stars. (They’ve even made a miniature star in their lab.) The facility uses 192 laser beams run by thousands of subsystems. Every day, scientists use NIF to perform physics experiments critical to national security and global energy research.

However, aging infrastructure, the gradual loss of legacy system expertise due to retiring team members, and failing equipment were hindering research. Random system failures and the lack of integrated expertise needed to fix them efficiently led to the loss of hundreds of hours of experimental time in 2024. LLNL needed a scalable, intelligent solution to efficiently troubleshoot equipment failures and retain dwindling legacy knowledge.

Shannon Ayers, division leader and technical manager at NIF, is part of the NIF Operations team determined to resolve these issues. They worked with LLNL’s AI Initiative to engage AWS Professional Services to integrate generative AI tools into NIF system searches. The team used Amazon SageMaker to create a chat-based assistant that uses Retrieval Augmented Generation (RAG) to perform intuitive searches of NIF’s vast problem log database. The tool creates step-by-step decision trees, which the staff used to access over 20 years’ worth of experiment and equipment recovery records instantly. Problems that previously delayed experiments by hours could now be solved in minutes, exponentially increasing their efficiency.

“AWS has helped to provide secure, scalable FedRAMP-compliant infrastructure, and advanced AI tools like SageMaker and semantic search have helped to accelerate our mission delivery through improved efficiency, reduce our technical debt by leveraging existing tools, and enhance innovation by bringing together thought leaders,” Ayers stated. “The collaboration with AWS has transformed our troubleshooting and is key to continued operational excellence at the National Ignition facility,” she concluded.

The accomplishment will help restore NIF’s capability to increase the number of experiments annually. In the future, the team intends to expand their predictive maintenance capabilities and explore operations management tools to support the agency’s mission-critical operations into 2040 and beyond.

USPTO realizes immediate boost in IT productivity using generative AI

The United States Patent and Trademark Office (USPTO) grants patents and registers trademarks to innovators. Due to rising demand, the agency needed to modernize its systems to deliver more digital services without increasing resource use. Like other agencies, the USPTO’s aging legacy infrastructure limited its ability to respond to the needs of the nation’s innovators.

Stephan Mitchev, former chief technology officer of the USPTO, had a vision for improving the efficiency of the department’s IT systems. He used AWS to deploy chat assistants and code assistants into developer workflows to help the team with tasks like unit testing, documentation, and reverse engineering of legacy code, which resulted in an immediate 20 percent productivity increase. As of May 2025, Mitchev had onboarded 27 teams into the new workflow, all of which reported increased positive outcomes.

Mitchev’s advice for teams is to start small with implementing the AI tools and measure the results. He emphasizes using AI to augment rather than replace existing staff and advises that teams measure and demonstrate return on investment (ROI) at each step to justify continued use.

Federal agencies are using AWS to modernize systems, increase efficiency, and improve their ability to serve the public. Connect with your account team to learn how AWS can improve the efficiency of your organization.