Government efficiency through AI-driven IT modernization with AFS and AWS partnership

Federal agencies are at a pivotal juncture in their technological evolution. Legacy systems, some dating back to the 1960s and 1970s, continue to power essential government services while facing a challenge: the retirement of the workforce that understands and maintains them. Rebecca Wetherly, director of worldwide public sector global system integrator partners at Amazon Web Services (AWS) recently sat down with Denise Zheng, chief AI officer at Accenture Federal Services (AFS), to discuss how artificial intelligence (AI) and modernization can help agencies navigate this transition while improving service delivery to citizens.

Legacy system challenge

During their conversation, Zheng highlighted several challenges facing federal agencies today, with legacy systems emerging as a central concern. The discussion revealed how these technological hurdles are creating unprecedented pressures on government service delivery and operational efficiency.

“Take the IRS, for example. They’re still running critical tax processing systems using programming languages that were developed back in the 1960s and 1970s,” Zheng explained. “The workforce that understands this language is aging.” This scenario isn’t unique to the IRS, it’s a challenge facing many federal agencies across the government landscape.

The impact of these outdated systems extends far beyond basic maintenance concerns. When Congress passes new legislation, implementing tax code changes can take 6 to 12 months simply because of the scarcity of specialists who can program in legacy languages. The complexity of these systems creates a cascade of challenges, where even simple updates to citizen-facing services require extensive coordination across multiple legacy platforms. As Zheng pointed out, “A simple update to a citizen-facing service requires coordination across five different legacy systems, and each has their own database, their own format.”

This technological debt is creating a growing challenge in government service delivery. As maintenance costs continue to rise, agencies struggle to find qualified personnel to support these aging systems. The resulting delays in implementing critical updates and service improvements are not just operational issues, they directly impact citizens’ access to essential government services. This results in citizens experiencing increasingly fragmented and delayed service delivery.

AI-driven modernization

The partnership between AWS and AFS offers agencies a path forward in addressing these legacy system challenges. Together, we are helping federal organizations tackle these issues through innovative solutions that combine AWS’s cloud and AI infrastructure with AFS’s government transformation expertise. Speaking about AFS’s experience, Zheng noted, “We’ve been able to take our commercial capabilities, federalize them, and help our government customers realize the efficiencies, quality improvements, and effectiveness that we’ve started to see scale across the commercial sector.”

A key innovation emerging from this partnership is Gen Wizard, a comprehensive modernization tool developed by Accenture with engineering input from AWS. Built and deployed on AWS GovCloud (US)  to meet federal security and compliance requirements, Gen Wizard leverages multiple AWS AI services to address legacy system challenges. The solution utilizes Amazon SageMaker for building, training, and deploying machine learning models that help analyze and understand legacy code patterns. Amazon Comprehend provides natural language processing capabilities to assist in documentation analysis and generation, while Amazon Bedrock offers access to foundation models that help modernize and transform legacy applications.

Gen Wizard accelerates new platform implementations, reducing the time and resources needed for system upgrades. It provides developers with AI-powered assistance for code migration and modernization tasks, helping agencies streamline their application portfolios and improve overall system efficiency. The tool speeds up data migration and modernization efforts, which is crucial for breaking down the information silos that hamper many agencies’ operations.

By leveraging these capabilities, federal agencies can begin to address their legacy system challenges more effectively. This approach improves operational efficiency and enhances service delivery to citizens, helping government organizations meet modern expectations for speed and accessibility of services.

Watch the full interview with Denise Zheng to learn how your agency can begin its modernization journey and discover how AWS and AFS are helping lead government digital transformation.

Realizing returns

When discussing the impact of AI-driven modernization on government resources, Zheng emphasized the significant potential for improvement. “On the cost side, we’re seeing that very manual processes, especially language-based tasks and workflows, which is a lot of workflows. As you can imagine, you can reduce costs and realize efficiencies of up to 40 percent,” she explained.

The return on investment (ROI) timeline is also compelling. Organizations typically begin to see positive returns within months of implementation, with more scaled returns materializing between 12 to 18 months. For example, agencies using Gen Wizard have reported significant reductions in development time. Amazon Bedrock’s foundation models help accelerate code modernization tasks that previously took months to complete. Amazon SageMaker’s machine learning capabilities, combined with Amazon Comprehend’s natural language processing, enable agencies to automate documentation and analysis tasks that traditionally required extensive manual effort.

Zheng emphasizes that achieving these results requires a thoughtful approach to AI implementation. “AI can automate these really time-intensive tasks and allow people to focus more on the higher value work. People need to always be in the loop,” she noted, highlighting the importance of augmenting rather than replacing human workers.

The key to maximizing ROI lies in leveraging the full spectrum of AI technology. This includes traditional optical character recognition (OCR), large language models (LLMs), and robotic process automation. By combining these technologies with a deep understanding of agency-specific workflows, agencies can achieve significant returns while maintaining the accuracy and reliability required for government operations.

Getting started

For agencies beginning their modernization journey, success lies in taking a measured, strategic approach. Zheng emphasized the importance of understanding the full spectrum of available technology rather than focusing solely on the latest trends: “We’re going through this phase right now where everybody is obsessed with agentic AI, myself included, but agentic AI requires maturity to make it deployable at scale, especially for agencies that operate very critical tasks in the national security or defense or even in the health space.”

The path to successful implementation starts with identifying specific workflows where AI can provide immediate value. This involves examining manual, time-intensive processes that could benefit from automation while ensuring human oversight remains in place. Agencies should focus on areas where they can achieve quick wins while building toward more comprehensive modernization goals.

Understanding the unique requirements of government operations is crucial. Zheng pointed out that agencies need to consider not just the technology itself, but how it integrates with existing processes and security requirements. This is particularly important in the federal space, where agencies must maintain strict compliance standards while improving citizens’ outcomes.

Conclusion

The modernization of federal legacy systems is a necessity for government agencies to meet the evolving needs of citizens and maintain operational efficiency. As Rebecca Wetherly’s discussion with Denise Zheng revealed, the challenges are significant, but so are the opportunities.

Through the strategic partnership between AWS and AFS, agencies now have access to innovative solutions like Gen Wizard that can transform legacy systems into modern, efficient platforms. With potential efficiency gains of up to 40 percent and the ability to see returns within months, AI-driven modernization offers a clear path forward for federal agencies.

The key to success lies in a measured, strategic approach that combines cutting-edge AI capabilities with a deep understanding of government operations and security requirements. By focusing on specific workflows, leveraging the full spectrum of AI technologies, and ensuring human oversight, agencies can navigate the complexities of modernization while maintaining the reliability and security crucial to their missions.

To dive deeper into how your agency can leverage the AWS and AFS partnership for legacy system modernization, we invite you to watch the technical deep dive video for detailed insights into Gen Wizard’s capabilities and implementation strategies.