How to Make AI Work for the Public Sector: From Data Swamp to AI Ready
The rapid growth of artificial intelligence (AI) in the public sector presents both tremendous opportunities and significant challenges. As government agencies increasingly adopt AI technologies, the need for a strategic approach to data management and infrastructure has never been more critical.
In the latest Tech Trends podcast episode How to Make AI Work for the Public Sector: From Data Swamp to AI-Ready Dillon Peterson, CEO of StandardData, and Jason Friend, CEO of C1 Gov, have outlined key considerations for public sector organizations to ensure they are truly "AI-ready" before diving into implementation.
Watch the Podcast
The AI Surge in Government: An Unprecedented Opportunity
Government interest in AI has surged in recent years, with agencies eager to leverage AI for everything from predictive analytics to process automation. However, this enthusiasm brings to light a key challenge: fragmented and siloed data. As Friend notes, “I have never seen the government embrace an emerging technology as quickly and as fully as they have with AI,” but this rapid adoption is often hindered by data that is scattered across different databases, formats, and even physical media.
The first step in AI readiness is understanding that data must be consolidated and organized before AI can be effectively utilized. Without a comprehensive data management strategy, AI efforts are likely to fall short, as the models and algorithms will struggle to function with disorganized or inaccessible data.
“The fundamental aspect of making AI work for an agency is the data -- the accessibility and the usability of that data.” Jason Friend, CEO, C1 Gov
The Challenge of Data Infrastructure
A significant issue facing government agencies is that many organizations simply don’t know what data they have. Data is often trapped in siloed systems and is not structured in a way that allows it to be easily discovered, searched, or leveraged for AI initiatives. The process of identifying, cataloging, and consolidating these data sources is foundational to any successful AI project.
Working in partnership, StandardData and C1 Gov provide AI assessments to help government organizations evaluate their current data environments and understand the extent of their data gaps. These assessments are not just theoretical; they include real-world proof-of-concept implementations, which help agencies visualize how AI can be applied to their specific challenges. These assessments are especially valuable for organizations with limited resources and open-source AI tools are increasingly becoming viable solutions for those looking to innovate on a budget.
Peterson underscores the importance of focusing on data infrastructure rather than solely on acquiring cutting-edge AI models. “AI models are becoming a commodity,” he says, and the real differentiator will be how well an organization can structure, clean, and manage its data. With a solid data infrastructure, AI models will have the fuel they need to deliver value.
Security, Compliance, and the Risks of Rushing In
While AI holds immense potential for government agencies, it also brings heightened concerns around security and compliance. Governments handle sensitive data, and public AI tools may not meet the stringent security standards required for such information. As Friend points out, traditional public tools are often inadequate when dealing with government-grade security needs, making it essential to implement secure, in-house AI solutions tailored to these unique challenges.
Another critical pitfall agencies must avoid is the temptation to rush into large-scale AI projects. Both Peterson and Friend emphasize that AI readiness is a gradual process that should be approached with caution. One common mistake is trying to do too much too quickly—agencies may want to launch high-impact AI initiatives right away, but without the proper data foundation, these projects are likely to fail.
For instance, Peterson shares the example of a federal archiving agency that began with a relatively small but impactful AI project using Optical Character Recognition (OCR) on millions of records. This step, though modest, provided a foundation for more advanced AI applications in the future. By starting small and building incrementally, the agency was able to ensure a successful transition into AI without overwhelming its systems or resources.
A Blueprint for AI Readiness
To help organizations navigate these challenges, Peterson and Friend advocate for a structured approach to AI readiness through the development of an "AI Readiness Blueprint." This blueprint starts with an essential first step: cataloging all available data sources within the organization. From there, the next phase is to identify the desired end state—what the organization hopes to achieve with AI—and then work backward to determine the necessary data sources to support that vision. The AI Assessment is a 6 week engagement that ends with a proof of concept, that organizations will be able to not only hear about the future, but see it in the form of a functional proof-of-concept.
“We can talk about pie in the sky AI ideas all day long, but unless it's actually matched to that organization's data and how that organization works, it remains a pie in the sky.” Dillon Peterson, CEO, StandardData
This approach allows organizations to see tangible results quickly, typically within six weeks. Rather than waiting months or years to validate the AI implementation, agencies can achieve concrete results in a matter of weeks, making the process both efficient and manageable.
As Friend aptly puts it, “Data is rapidly becoming an agency's most valuable asset.” In today’s data-driven landscape, organizations that prioritize data consolidation and standardization will be better positioned to leverage AI successfully.
Data is the New Oil
AI is transforming the way government organizations operate, but this transformation cannot happen without proper preparation. By focusing on data consolidation, security, and a structured implementation process, agencies can lay the groundwork for successful AI adoption. For those looking to get started, StandardData and C1 Gov’s affordable AI assessments offer a practical way to assess current data capabilities and develop a roadmap for AI integration.
Data is the new oil, as Peterson notes—just as oil needs refining to unlock its full potential, so too does data need structuring and standardization to maximize its value. By taking a systematic approach to AI readiness, government organizations can unlock the true potential of AI and drive meaningful digital transformation.
