If there's only one takeaway from the recent Pryon Government AI Forum, it's that federal government organizations are chomping at the bit to implement generative AI (GenAI) in a responsible, ethical, and impactful way. From the military to civilian use cases, there are many ways GenAI can help enable decision advantage, improve citizen services, and streamline information delivery.

We wanted to help illuminate where federal organizations should focus their efforts, what to keep in mind, and how to keep the momentum going until a GenAI application is in place and producing results. To that end, we brought together six of Pryon's leading federal experts and asked them the questions government officials have top of mind. Keep reading to see what each of them had to say.

Identifying the best use cases for AI in the federal government

QUESTION

What government-specific challenges can AI best address?
‍

‍

Mike DiCarlo, Federal Sales Engineer, shared the following tidbits:

• Government agencies are increasingly turning to AI to address key challenges, including knowledge friction, productivity, accuracy, and user experience.
• A few examples include:
  • A federal agency uses Pryon RAG Suite to help its analysts quickly research and analyze content spread across disparate content sources. The analysts can quickly get answers, attributed to the source. The agency also enables an easy, reliable question-and-answer experience for citizens instead of forcing the public to look through the agency’s website.
  • A military organization leverages Pryon’s AI in both public-facing chatbots and internal processes to deliver precise responses to technical queries from manuals and flight operation documents in situations when accuracy is vital.
  • Another organization uses AI to efficiently manage complex policies, ensuring adherence to the risk management framework.

‍

"Pryon can address what's called knowledge friction, or that disconnect between where content resides and what end users need of it. Those end users include war fighters, internal employees at an agency, and taxpayers."

‍

Allaying concerns about AI

QUESTION

What concerns government leaders and acquisition teams when it comes to AI?

‍

Here are some quick-hit points from John Patterson, Federal Sales Engineer:

• Government leaders often fear AI could lead to scenarios where technology replaces human jobs, echoing concerns depicted in science fiction.

• These leaders should feel assured that the objective is to augment human capabilities, not replace them, by automating routine tasks and freeing up time for more critical work.

• Retrieval-Augmented Generation (RAG) is a pivotal solution for delivering immediate access to trusted information, enhancing productivity, and supporting successful mission outcomes.

‍
‍

"The number one concern we're hearing over and over again is that [government leaders] think AI is going to be like the Terminator, with SkyNet. AI will be like a magic button that's going to replace their entire human workforce and lead to World War 3...but we're not trying to replace the human with a button."

‍

‍

QUESTION

How can federal organizations ensure the security and privacy of sensitive content?

‍

‍

John shared the following insights:

• Ensuring AI security and privacy hinges on secure implementation and tailored solutions to fit the scale of deployment.
• Some RAG vendors, like Pryon, offer solutions that can be hosted entirely on-premises, air-gapped, or within a secure government cloud environment.
• These secure setups ensure that sensitive questions and content remain protected, preventing exposure to the internet, foreign adversaries, or private entities.

READ MORE: Generative AI for Government: Practical Applications and Implementation Guide

‍

Going from idea to implementation

QUESTION

What are some examples of innovative or high-impact AI implementations you’ve seen in federal organizations?

‍

‍

Here's what Keith Howard, Lead Federal Solutions Manager, had to offer:

• Generative AI and RAG are transforming access to unstructured information, allowing federal organizations to quickly retrieve data from documents like PDFs and Word files using natural language queries.
• Public-facing applications, such as chatbots, have been successfully deployed by numerous government organizations to enable the efficient querying of extensive repositories, including DoD instructions and Air Force publications.
• Advanced tools are being developed for automating report generation and creating up-to-date PowerPoint decks, enhancing the ability to handle time-sensitive information swiftly.
• Integrating AI with solutions like ServiceNow introduces automation capabilities, facilitating workflows and rule-based actions.

QUESTION

Once a use case has been defined, what timeline can federal organizations expect to go from idea to implementation?

‍

‍

According to Keith:

• Identifying user needs and workflows is a critical first step to implementing AI solutions, as this helps determine the appropriate technology and approach.
• Some RAG systems, like Pryon RAG Suite, can be deployed rapidly — they often take only a few hours to connect with existing data sources.
• Fine-tuning large language models can be time-consuming and resource-intensive, requiring dedicated teams and significant effort, making it less ideal for quick implementations.
• Tailoring technology choices to specific use cases ensures more efficient implementation, aligning solutions with the unique requirements and constraints of federal organizations.
  ‍

‍

"It all goes back to the user. Identify what the use case is...that will drive your implementation time."

‍

QUESTION

Are there lessons that the public sector can learn from the private sector when adopting/implementing AI solutions?  

‍

‍

Vince Berry, Services Consultant, provided this advice:

• The private sector's emphasis on scalability and agility allows for quick adaptation to changing demands and technological advancements. The public sector can embrace this mindset to improve responsiveness and competitiveness.
• Cross-functional collaboration is pivotal in commercial enterprises by integrating AI across various business processes, encouraging teamwork between AI teams, business units, and stakeholders, and by aligning initiatives with overarching organizational goals.

• By adopting these strategies, federal government organizations can better position themselves to leverage AI technologies effectively, maximizing their impact and ensuring alignment with mission-critical objectives.
  ‍

QUESTION

How should federal organizations go about acquiring funding for new AI solutions?  

‍

‍

Here's what Stacy Brown, Director of Defense Programs, recommended:

• Federal organizations can acquire AI funding by educating and guiding government leaders and officials on the value and safety of AI, ensuring its prioritization in budget planning and alignment with legislative processes.
• Leveraging existing AI strategies, such as those released by the Department of Defense and the Department of State, can help align AI initiatives with government policy and funding priorities, ensuring a strategic fit with national objectives.
• Unfunded requests and research and development (R&D) funds provide alternative avenues for securing resources, even if AI isn't initially included in designated budget lines.
• Emphasizing the potential impact of AI on transforming government operations and future-proofing can reinforce the case for AI investment.

‍

Assessing impact and getting started

QUESTION

How can federal organizations assess the impact of implementing new technologies like GenAI?

‍

‍

Here are some tidbits from John:

• Federal organizations prioritize mission impact over traditional ROI when assessing the implementation of new technologies like generative AI. They want to ensure these solutions enhance their ability to achieve organizational goals effectively.
• Successful technology integration requires embedding these tools into existing workflows, minimizing disruption and ensuring that analysts can work more efficiently without additional complexity.
• The focus should be on delivering targeted solutions that directly address mission-specific needs. This reduces time to implementation and fosters improvements in operational efficiency and mission success.

‍

"If [an AI application] is just noise, if it's just adding another tab that an analyst has to have open, it's not helping."

‍

QUESTION

How should federal organizations plan to scale a GenAI deployment and decide where to take it?

‍

‍

Some wise advice from Jeremy Hogg, VP of Federal Sales:

• Start by clearly defining the core use case and understanding the specific problem to be solved. Make sure all relevant stakeholders are involved from the outset to align on the business value and objectives.
• Ensure robust infrastructure and security measures are in place, since GenAI deployments require substantial computational resources and secure platforms that comply with federal standards.
• Leverage content sources effectively by applying technologies like RAG to enhance data accessibility, ensuring that content is reliable and trusted by end users.
• Focus on achieving early wins to build trust and demonstrate value to initial users. This will turn them into advocates for the platform, which helps lay the groundwork for broader adoption and scaling.
• Engage in an iterative process of continuous evaluation and feedback collection to refine the deployment, ensuring it consistently meets the evolving needs and challenges identified at the project's inception.
  ‍

"Once those three pillars are defined — the use case, infrastructure, and security — you can start moving on to, well, how will this project actually scale out?"

‍

Wrap up

While the road from idea to implementation isn’t an easy one to take, especially in environments as sensitive and risk-averse as many federal organizations, the benefits of deploying AI can be profound. Military and civilian organizations can reap many rewards from implementing AI solutions, such as chatbots and automated report generators.

‍

But doing it right requires organizations to:

• Prioritize accuracy, security, speed, and scalability
• Involve all relevant stakeholders from the start
• Deliver targeted solutions that augment human intelligence
• Deploy solutions that can quickly adapt to new needs

‍

Learn more about AI for federal government organizations through the following resources:

• Blog post: Generative AI for Government: Practical Applications and Implementation Guide‍
• Guide: Generative AI for Federal Government‍
• Blog post: Key Insights from the Pryon Government AI Forum