Following the release of ICF and Wakefield research’s new annual report, Data and AI: The top trends shaping government in 2024, OS AI sat down with Chief Technology Officer Kyle Tuberson to better understand some of the current challenges and opportunities for AI within the federal government, including use case prioritization, defining a North Star, and how ICF views the technology ecosystem.
AI and Use Case Prioritization
With the growing interest in and adoption of AI, use case prioritization becomes a key consideration. Among the potential benefits is citizen engagement, built upon journey mapping and understanding value streams. Another is around helping government align and mobilize its different capabilities. Finally, there are more narrow use cases as might exist within finance or HR departments.
“There is so much promise, but this opportunity also requires clearly defined use cases to ensure outcomes are clear and being met,” said Tuberson. “There are quick wins, such as speeding up the time to complete a task, but then there are the bigger picture items, such as those related to re-envisioning entire end-to-end business processes. For example, we’re working with CDC to identify whether generative AI can help improve public health outcomes. Instead of devising a discrete solution for the agency, we took a methodical, exploratory approach to identify the best pilot use cases, from automating code to streamlining data surveillance. As a result, our human-centered AI approach helped CDC identify real-world workforce-augmenting applications.”
When prioritizing use cases, it’s critical to take a close look at the underlying data. Building on a foundation of curated data that is findable, accessible, interoperable, and free from harmful biases can uncover game-changing AI opportunities. “Many organizations are fast getting out of the gate with the technology,” said Tuberson, “but then they run into data hurdles and challenges the AI uncovers.”
Data Collection vs Data Use
Noting that fundamental data challenges have been in existence and known about for some time, now is the time to understand what they are, how to mitigate, and how to extract value. “How do we leverage Generative AI technology to classify sentiment and regulatory comments?” said Tuberson. “It’s about making sure that you know we understand where the data is and how to extract the value out of it.”
Looking from the lens of those who use data, and those who collect it, each will say their area is the more challenging, creating a dichotomy, and a further block to understanding where the challenges reside. “Ultimately, it is less about the tool side than the people side, as in how they collect and use data. The tools and all that go with it are easy enough to deploy.”
ICF’s recent report noted that many data projects fail when the domain experts are not included in modernization efforts. Subject matter, or domain, experts bring extensive knowledge and skills within a specific area or industry crucial to a mission’s success. Without them, the richness of the outcomes that can be driven are not both front of mind. According to Tuberson: “We found that 88% of federal IT professionals say that their projects fail without a domain expert on it. But despite this understanding, 32% admitted that their agency’s digital modernization efforts did not include a domain expert. Clearly the solution is understood but yet, it isn’t happening.”
Other challenges noted in the report include governance, privacy issues, and a lack of established rules to allow decision making across the enterprise. “Within ICF, AI enablement is a corporate-wide sport bringing in administration and legal, HR, the CIO, and those using AI day to day,” said Tuberson. “It’s important to start with a set of decision rights that have been communicated. People have to be empowered to make decisions that make sense, and to know what may need elevating. Moreso than any technology I’ve ever experienced in my 25 years, AI really requires everybody to come together to establish a vision.”
Innovative and Disruptive
Understanding that AI is as innovative and as disruptive as anything that has been seen in a long time, ICF has established AI express lanes that empower the workforce to use and experiment with AI without bogging the organization down. This allows modernization and transformation to happen not only quickly and naturally, but also responsibly and ethically.
“It takes an effort to mobilize leadership,” said Tuberson. “Maybe you’ll determine that 10% of what you do needs heavy governance, but for the other 90%, you can empower the workforce. In that case, give them clear decision paths so that they can go from concept idea to value as quickly as they can.”
Whether it is for government, or on the industry side, this philosophy requires clear objectives and an understanding of the value AI will bring. Then, it is up to organizations to prove their path using use cases and small pilot projects that provide demonstrable success.
North Star and the AI Opportunity
The point of a pilot, says Tuberson, is for challenges to surface that can then become the North Star of an organization, or an area that needs focus, a vision and strategy. Using AI, workers can be cleared to focus on the richer visions, the bigger pictures. “I love data, but now I have this ability with generative AI to look for the needles in the haystack. I can now start to explore and experiment with unstructured data that’s harder to extract value out of. AI unlocks that data, but it still requires human creativity to untap it and to extract the value.”
It will be no surprise that 38% of mission leaders cite cultural resistance as the reason their agency’s digital modernization efforts fail. Overcoming this can require “gifts and goodies,” or tangible wins that go beyond just talk, and enable decision makers to see the value of these efforts, encouraging them to get on board.
“The technology is an afterthought when decision makers focus on the outcome,” said Tuberson. “For the best results, you need to look through their lens and understand that much of the workforce will resist change in the early days.” He added: “It may be necessary to identify the early adopters in any change curve and tap them to help drive a tech-forward culture.”
Ultimately, success means getting better at measuring outcomes and crystalizing the organization’s North Star. It’s essential to be clear on the objectives targeted so that, across the organization, forward movement is understood. Tuberson notes that, often, the struggle relates to a focus on outputs, rather than outcomes. “When you set your North Star on outcomes, understand movement will be measured, then you can mobilize the team, empower decisions at the edge, and then let everyone go do what they’re good at.”
Being Part of the Technology Ecosystem
While a key player in driving digital modernization, especially among the federal government, ICF exists as part of an ecosystem, bolstered by capabilities of larger companies, and gaining innovation and new perspective from smaller ones. “When you expand beyond the everyday use of AI, you start to get into cloud engineering, automation, and advanced AI skills,” said Tuberson. “We’re always looking for partners, those who have worked their whole careers in these areas and really understand the potential outcomes.”
On the technology side, ICF partners with Salesforce, Appian, AWS, Google, Microsoft, Databricks, Snowflake, and ServiceNow to name a few, digital modernization powerhouses who are driving innovative technology into the market. Said Tuberson: “Plugging our domain expertise in mission areas, we all become part of this ecosystem working together, which is really important for us.”
About Kyle Tuberson
Kyle Tuberson is ICF’s chief technology officer, leading the company’s technology, data management, and product innovation efforts. He both defines and manages key strategic initiatives that help employees and clients embrace technology as a game-changing solution that will achieve better, faster outcomes for individuals and communities worldwide. Leveraging 20+ years of technology consulting experience, Kyle’s team leads provides technology and IT modernization services to government agencies and in industries such as climate, public health, disaster management, and energy. These services include data modernization, cloud computing, artificial intelligence, internet of things (IoT), geospatial analytics, data visualization, and more. Kyle leverages AI/GenAI to create solutions that boost workforce efficiency, optimize service delivery, and accelerate innovation. He works with agency leaders to get the most value out of AI/Gen AI by creating strong data foundations and by bringing in experts to interact with large language models in sophisticated, domain-driven ways.
In 2023, Kyle was named a Northern Virginia Technology Council Tech 100 Executive honoree and a WashingtonExec Top Chief Technology Exec to Watch. He is also a 2018 Federal 100 award winner and was awarded the CIO University Certificate in Federal Executive Competencies by the Federal CIO Council.
Want to get involved with OS AI? - A small number of Sponsorship Opportunities are now available here. Starting at $500.