From Data Strategy to Artificial Intelligence Strategy: The Golden Hexagon Pathway
Abstract
The Artificial Intelligence (AI) adoption spectrum across the United States government is very uneven. While most agencies have implemented simple AI-based applications such as automating simple workloads, developing application specific dashboards, or experimenting with simple machine learning predictive models, others are exploring cutting-edge AI-based solutions for near real-time intrusion and detection, recognizing objects in video data, or performing predictive maintenance. The problem is in the middle of the adoption spectrum. This paper introduces the DATICS framework or the Golden Hexagon, an AI Framework to guide Federal agencies in their AI journey.
The Data Strategy
The data strategy is the vision and roadmap for developing and maturing data-driven enterprise capabilities for organizing, governing, and deploying data assets and capabilities. The objective is to harness and align data and analytics activities towards organizational goals. The data strategy is essential in adopting and scaling Artificial Intelligence (AI) to drive significant business value, steer operational efficiency and effectiveness, and optimize return on investment. It guides the development of a portfolio of data supporting and management tools and techniques to ensure the delivery of consistent and high-quality data, which is the foundation for successful, productized AI.
Towards a long-term vision for modernizing the federal government, the United States Federal Data Strategy (FDS) describes a 10-year vision for how the Federal Government will leverage data to support agency missions and serve the public. It provides a common set of data principles and best practices in guiding agencies to establish a data strategy and accelerate its implementation. The components of FDS are shown in Figure 1 (The Federal Data Strategy, 2020).
Figure 1. The Federal Data Strategy Framework
Data Maturity
While FDS is central to the vision for modernizing the Federal Government and driving data innovation, data maturity capabilities are at different levels in the AI adoption spectrum. One of the goals of FDS is to empower agencies to develop a mature data asset management environment to leverage data as a strategic asset. An essential part of that journey is to evaluate, improve, and grow the capability of using data to achieve mission outcomes.
Data maturity can be defined as the ability to assess the level of maturity of an organization’s data management and undertake continuous improvement in data management, governance, and analysis. Data maturity and AI are critical in addressing national challenges and empowering Federal agencies to engage in the process of shaping enterprise-level data strategies to address national priorities and support evidence-based decision making, while efficiently and effectively delivering mission outcomes (Chraibi, 2020).
There are several data maturity models available. Figure 2 shows the federal government data maturity model (Data Cabinet, 2018). It is a general representation of data and analytical models that provides several dimensions to help agencies establish a data maturity gap analysis at a high level in order to support the data strategy towards data maturity. Data maturity models are helpful in the process of assessing the capabilities of an agency in key data related areas, as well as charting out milestones.
Figure 2. The Federal Government Data Maturity Model
The many challenges that impede the adoption of AI and emerging technologies include the lack of a robust strategy, the underwhelming support for the modernization of systems, the tepid development of data-driven platforms and applications, and the scarcity of the workforce to implement, monitor, and leverage data-driven solutions.
AI Adoption Challenges
AI capabilities are to be leveraged at the technical, organizational, and operational levels. There are several challenges to AI adoption:
· Data Challenge: One of the most critical components of any operational entity is its data. In most agencies, data may be difficult to access or is of poor quality due to lack of proper data collection, management, and governance. There is a also a lack of a unified data management platform with tools and techniques for collecting, securing, analyzing, disseminating, and managing data-driven solutions.
· Innovation Challenge: Driving the transformation of the federal enterprise through innovation is essential to achieve mission outcome. However, innovation is a hard journey. It requires from any organization to transform itself and change its processes, services, business models while integrating risk and change management efforts. To succeed, a federal agency should proceed with an operating framework that enables it to explore innovative projects with tangible, immediate impact congruent with its national mission. Furthermore, innovation often involves failures as well as successes. Often, it happened in the margins while true digital transformation needs to happen at the core.
· Integration Challenge: Agencies have been using data to support specific business applications in siloed environments. This configuration impedes the leverage of data value and its full utilization for cross-cutting projects.
· Scarce Talent: The short supply of AI talent and advocacy precludes agencies from focusing on building diverse teams to cover the breadth and depth of AI projects lifecycle. The need is exacerbated as additional skills such as possessing business and Information Technology (IT) acumen, as well as mastering project management and communication skills are fundamental for AI success.
· Modern IT infrastructure: Many agencies continue using outdated systems presenting updates, security, and integration risks. Efficiently developing, implementing, and monitoring AI technologies requires modern and effective infrastructures.
· Change Management: Change is difficult when it requires from any organization to transform itself and change its processes, services, business models while integrating risk and change management efforts.
Succeeding in this effort requires a robust, healthy, transparent ecosystem of diverse data sources, human and infrastructure resources, and AI capabilities to support the entire business process from the genesis of the data to the experience of the end user.
However, there are several risk factors such as effectively managing change and disruption, readying the workforce, developing and deploying AI-based complex products and services, leveraging an ethical and compliance framework, and ensuring data privacy and quality.
The AI Golden Hexagon
While the United States Federal Data Strategy is a central component in promoting the adoption of Artificial Intelligence (AI) within the federal public sector, there is a need for a roadmap for AI success. I propose here an AI framework that includes six pillars/components that I coined the AI Golden Hexagon to effectively guide data-driven digital transformation (see Figure 3).
Figure 3. The AI Golden Hexagon
The six pillars include:
1. Data: The most value comes from being able to collect, aggregate and correlate information from different kinds of systems. To leverage the necessary data to support operations and mission-critical applications, access to large volumes of high quality data at high velocity with different data types and flows is crucial.
2. Data Strategy and Policies: Capabilities and governance policies to support the ingestion, discovery, curation and quality of data, as well as to ensure data security, privacy, compliance.
3. Analytics Platform: Hardware, software and tools framework that enable and accelerate enterprise-grade AI projects. It should support the full Machine Learning/AI lifecycle from data genesis to model production, deployment, and monitoring.
4. Modern IT Infrastructure: A flexible, secure, powerful infrastructure where agencies are able to maximize secure use of cloud computing and effectively support the agency applications.
5. Talent: A diverse mix of talent to translate mission needs into solution requirements, including the structure, skills, and processes required to execute AI projects at scale.
6. Data-Driven Culture: Agencies must establish the organizational capability, including the support for a data-driven culture, a framework for responsible and ethical data-driven decisions, a system of governance and change management, all aligned with the agency mission and priorities.
Summary
Establishing a comprehensive data strategy is fundamental to lay out the foundations for a successful and sustainable AI strategy. Building on the data strategy, this paper proposes six pillars of an effective AI strategy, coined the AI golden hexagon, to develop, scale, and maintain successful and cost-effective AI.
References
Overview — Federal Data Strategy. (2020). Retrieved from Strategy.data.gov:
https://strategy.data.gov/overview/
The Federal Government Data Maturity Model. (2018). Retrieved from https://my.usgs.gov/confluence/download/attachments/624464994/Federal%20Government%20Data%20Maturity%20Model.pdf?api=v2
CHRAIBI, CHAKIB (2020, May 27). Artificial Intelligence in the U.S. Government: A Mission-Oriented Journey from Dark Data to Open Data and Data Innovation. [Article]. Medium.Com/FirmAI. https://medium.com/firmai/artificial-intelligence-in-the-u-s-fcf594ee6976
