Building AI Competency

Last month, I wrapped up my two-part essay on business transformation12 with a conclusion: companies that started early are no longer “transforming”, they are in a state of continuous evolution. Google remains a prime example of how even pioneers must perpetually adapt to stay ahead. Most C-level executives recognize the imperative, yet many struggle to match the velocity at which pioneers like Google, Walmart, and JPMorgan Chase have pivoted in just the last twelve months.

In 2026, the challenge for CEOs is no longer fighting the inertia of “getting started.” The challenge is architecting a culture of continuous evolution, one that sustains momentum long after the initial transformation push.

Coincidentally, the January–February 2026 issue of Harvard Business Review led with “Rethinking Nonstop Transformation.”3 The cover story, by Darrell Rigby and Zach First, challenges the modern corporate obsession with “radical reinvention.” They argue that constant cycles of massive, disruptive change are counterproductive and exhausting. Instead, they advocate for steady, proactive adaptation. The issue’s mantra is powerful:

“The best way to manage transformations is to minimize the need for them.” True progress comes from building a resilient, “regenerative” business model that evolves alongside the market, rather than one that must be broken and rebuilt every three years.

The shift from radical upheaval toward sustainable evolution has fundamentally reshaped my practice at Kai Roses. I have moved away from the traditional, linear path of discovery phases, maturity assessments, and isolated pilots. Instead, I focus exclusively on helping companies build AI Competencies as the primary path to value. This is achieved through accelerated daily and weekly iterations, where we develop multiple model-based solutions simultaneously to surface gaps and opportunities with speed.

Given that AI will radically change how businesses are built and operated, we must first understand the true nature of the organizations we are evolving. This essay, the first in a 12-part series introduces the foundational concept of Competency through a simplified industry lens. In the coming months, we will dive deeper into the core competencies of modern industries to understand how they might leverage AI to build new competitive moats. Along the way, we will define exactly what AI Competencies are and, more importantly, how to develop them within your own organization.

Understanding Competency

In 1990, C.K. Prahalad and Gary Hamel defined core competencies as the “distinctive, often intangible combinations of knowledge, skills, and abilities” that form a company’s unique strengths4.

They famously compared a corporation to a tree: the roots are the core competencies, the trunk represents core products, the branches are business units, and the leaves are the end products consumers buy.

Their research remains as groundbreaking today as it was in the 90s; I highly recommend a deep study of their work, as it fundamentally shifts how one views the structure of a modern business. I am particularly grateful to Professor Vijay Gurbaxani for directing me to this paper during my own research into AI Competency as a Senior Industry Fellow at the UCI Center for Digital Transformation.

To apply the core competency framework in today’s business landscape, consider a simplified example: a bottled water company. Today, such a company likely relies on traditional software to manage its network of sourcing, bottling, and distribution. However, they have yet to move toward AI-fueled decisioning across their value chain:

• The Roots (Core Competency): Expertise in water sourcing, logistics, and high-volume bottling.

• The Trunk (Core Product): High-quality spring water.

• The Branches (Business Units): Supply Chain, Marketing, Finance, IT, HR, etc.

• The Leaves (End Product): Bottled water available at retail.

Now, consider how this same company might operate to compete in an AI economy, one where the creation, capture and distribution of value are fueled by AI:

• At the Roots: AI predictively optimizes water table sourcing and replenishment.

• In the Branches: AI streamlines bottling logistics via real-time demand sensing or hyper-personalizes marketing at scale.

• On the Leaves: An AI-integrated “smart bottle” that tracks consumer hydration in real-time.

If this company simply buys a chatbot (AI on the leaves), competitors will copy it within weeks; at best, it becomes a PR statement forgotten within days. If they buy AI for logistics (AI in the branches) but fail to invest in the Roots, the tech will fail to integrate into the end-user experience. This results in only short-term efficiency gains via expensive siloed AI that do not translate to sustainable differentiated moat. This is the danger of surface level AI: it offers no moat.

The bottleneck (pun intended) isn’t the availability of AI tools and vendors, it is the lack of internal knowledge, skills, abilities and infrastructure (AI at the roots) to move data at the speed and volume necessary to monetize value and deliver true market differentiation.

From Narrow Roads to an Eight-lane Freeway

AI thrives on two key levers that legacy software infrastructure was never designed to handle:

1. Speed: Information flow must move faster than any legacy system.

2. Context: Systems must process a much larger quantity of data on an ongoing basis.

Previously, businesses operated on narrow roads with high-deterministic software, input A always led to output B, ambiguities were managed through business processes. Now, these business will need an eight-lane freeway for data to flow, providing probabilistic decision support (where the system predicts the best “A” based on shifting context) across their entire value chain.

Most IT departments have had over two decades to build the software equivalent of these competencies. To compete in 2026, companies face the challenge of building the AI equivalent in a fraction of that time. This requires more than just hiring; it requires a thoughtful closing of expertise gaps, upgrading existing skills, and expanding abilities across both IT and core business functions.

There is another strategic friction: you need the competency to win, but you need a win to fund the competency. Funding for AI often originates in sales or commercial teams eager for immediate action. Conversely, securing a budget purely for “competency building” is a difficult sell in a results-oriented boardroom.

The most pragmatic way to break this cycle is to weave competency building directly into your active projects. In the early phases, there should be a heavy emphasis on visible value creation, but with a disciplined allocation: 25% to 30% of project funding must be diverted toward foundational talent, governance, and infrastructure. Even with a modest portfolio of two or three projects funded at $250K to $300K each, it is possible to start developing AI competency. This approach treats these first “sprints” as a dual-purpose investment:

1. Direct Value: Solving a specific commercial problem.

2. Infrastructure ROI: Building the “Eight-Lane Freeway” that makes all future projects significantly cheaper and faster to deliver.

Without investment in AI competency, it will be impossible to evolve from a software-backed deterministic operating system to an AI-backed probabilistic one.

Conclusion

Over the past year, I’ve found that most clients need an AI advisor, a project leader, a hands-on technical expert, and a governance lead, all at once. They require a thought partner who can operate at a 30,000-foot strategic level while acting as an operator on the ground. My current practice reflects this reality, drawing on a 27-year career in AI and data science as a hands-on executive comfortable with high-velocity, complex model development and delivery.

I now limit my practice to just two companies at any given time, requiring a 90-to-180-day retainer. During this window, we develop and iterate on 5 to 10 model-based solutions, an approach far more effective than hiring a vendor for a single, expensive “black box” model that may prove impossible to integrate into an existing software stack.

In 90 days, we can uncover immediate ROI, detect critical skill gaps, and begin prioritizing the most critical internal “roots” required for a differentiated competitive moat. This model delivers three primary outcomes:

• Cost Efficiency: Avoiding massive investments in models without a clear implementation plan.

• Velocity: Saving months of development time through rapid, parallel iterations.

• Strategic Clarity: Building organizational confidence in choosing the specific AI applications and competencies the business must prioritize.

This method follows a non-linear path where discovery and assessment happen in parallel with real development. When a core competency is missing, traditional discovery is flawed; you cannot interview people about a paradigm they have never experienced. For mid-market firms and the Fortune 500 alike, AI is not a software add-on; it is a core competency that is currently missing from most organizations.

Next Month: I’ll explore how core competencies vary across different business domains and why the "competency gap" will be the primary factor that separates market leaders from laggards in the AI economy.

About the Author: Meghna Sinha is Chief AI Officer and Co-founder of Kai Roses, Inc. Kai Roses was founded to solve for the net-positive impact of AI on culture and commerce. Our consulting service is focused on helping companies build AI Competency.

Contact Meghna at kairoses.com to discuss how to navigate your company’s evolution in the AI economy of 2026.

Thank you for reading my essay. If you found it insightful, please comment, share, like.