Analytics to AI: Advancing the Journey

Federal agencies recognize the power of data to help inform better decisions and drive better mission outcomes. Many have made real progress in the journey to become data-driven organizations—and are preparing to integrate artificial intelligence (AI) to support advanced analytics. AI can help address daunting backlogs, workforce capacity limitations, high transaction volumes and other challenges that often hinder federal agencies’ ability to deliver mission outcomes. But AI succeeds only with a strong foundation.

Laying the groundwork

A common first step is getting a firm handle on an agency’s programs and program data, and strategic plans for mission areas—and identifying where and how analytics can help solve issues. Progress may also include creating an agency-specific lexicon to support data classification and contextual sentiment analysis, as well as standing up an Analytics Center of Excellence to serve as a focal point for capturing expertise and disseminating best practices throughout the agency.

From there, an agency can implement the resources needed to arm every level of the organization with actionable analytics: a Master Data Management strategy, simple dashboards visualizing the existing data for users, and self-service data capabilities for internal stakeholders. These help to shine a light on your data so you know where it needs to be corrected or adjusted.

These are areas many Agencies, like USDA, are addressing now to prepare for future uses of data, including AI. After all, clean data is a critical foundation for any organization seeking to use AI and machine learning to support advanced analytics. With that as a foundation, your agency can make the critical step from simply attaining a more accurate, complete view of what’s already happened to building a data-driven view of what’s likely to happen in the future.

Creating an AI strategy

Making the leap to predictive analytics is a classic case of “walk before you run.” If your agency has tackled the upfront work of getting your data in good shape, it’s time to consider how and where to apply AI. While it might be tempting to plunge into point solutions, a better approach is to craft an AI strategy the covers the key strategic, operating model and tactical items critical to sustainable success.

To get started, think through questions in these key areas:

Strategic

• How will AI affect our constituents, and what does that mean for the organization?
• Where is our agency most vulnerable to AI-induced disruption?
• How will AI’s changes in human-machine interactions affect our workforce?

Operational

• How does our workforce need to change to enable AI?
• Where are our talent deficits? How do we attract or build talent to support AI?
• How will AI be governed?
• How do our business and IT operating models need to change to enable AI over the near and long term?
• How do we scale AI across the agency?

Tactical

• Which parts of the agency are best suited to leverage AI over time?
• Which specific opportunities for using AI have the highest ROI?
• Which standalone AI proof-points would be good for AI pilots?
• Which types of AI are likely to be of most benefit to the agency and over what timeframe?
• What development process will we use to create and deploy AI solutions?
• How does the agency’s technical architecture need to change to support AI?

From there, you can begin to build real momentum with AI. Start identifying specific use cases for AI-enabled advanced analytics. Application and request processing, financial fraud and error detection, customer dashboards and career personal assistants are just a few of the potential applications to improve customer engagement and employee experience.

The journey from analytics to AI won’t happen overnight, but the destination—delivering better mission outcomes—is worth the effort.