As organizations attempt to keep pace with the rapidly evolving landscapes of artificial intelligence and data analytics, many are developing AI centers of excellence (COEs). The goals are to harness the power of AI to create business value, to drive innovation, and to stay ahead of the competition.
AI COEs follow the traditional COE model, though obviously focused on the role of AI in business strategies and in corporate culture. They draw the top talent from throughout an organization that can champion how AI can be interwoven into existing processes to create new efficiencies.
Consider the example of ECS, a leading provider of cloud, cybersecurity, AI, machine learning, and IT modernization services in Fairfax, Virginia.
“Our data and AI center of excellence was established to advance our company’s data and AI culture, practice, products, partnerships, and social eminence,” explains Rick Torzynski, senior data and AI engineer and product architect for Atlas Graph at the company. “With a vision to scope, prioritize, induct, govern, and integrate data and AI opportunities, our COE aims to align our capabilities with the strategic vision and operational needs of our customers.”
By creating a dedicated COE for AI and data analytics, Torzynski says the company can leverage the expertise of over 200 data professionals — 50% of whom hold PhD or master’s degrees — to drive innovation and solve complex problems.
“Our COE serves as a community of practitioners, providing a platform for knowledge sharing, expertise exchange, and best practice development,” Torzynski says.
Investments for the Future
As an AI leader, practitioner, and researcher, Adnan Masood has long maintained that an AI COE defines a strategic inflection point in how organizations can future-proof their core competencies.
The chief AI architect at UST, Masood says his organization established its COE to address a classic demand: “We needed adaptive leadership and a single source of truth for data-driven strategy, from dynamic risk management to value chain optimization,” Masood explains. “That has been our springboard to scalable solutions and sustained momentum. Since then, I have worked with various client organizations, helping them establish and run their respective COEs.”
UST, formerly known as UST Global, is a provider of digital technology and IT transformation services based in Aliso Viejo, California. But AI COEs are hardly limited to technology companies. Among non-tech organizations, Masood says he has watched COEs generate quick wins — such as reducing fraud loss by double digits or serving as a catalyst for deeper initiatives such as advanced customer lifetime value analytics.
When an AI COE is designed and staffed effectively, executives can hope to witness strategic cohesion when AI-driven insights support critical-path decisions, aligning people, platforms, and cultural capital, Masood says.
“AI COEs help build cross-functional synergy by merging data scientists, domain experts, and finance leaders,” Masood explains. “That cross-pollination strengthens institutional memory while mitigating the strategic ambiguity that so often stymies new tech deployments.”
Potential Benefits from an AI COE
The benefits of an AI COE can be many and go far beyond technology advancements, Torzynski says. At ECS, they include:
- Talent development: “Our COE provides opportunities for employees to grow professionally, even in areas outside of their primary job responsibilities. Regular town hall meetings encourage employees to explore various COEs, fostering a culture of continuous learning and development,” Torzynski says.
- Knowledge sharing: “Our COE serves as a community of practitioners, holding regular meetings and events to share knowledge, expertise, and best practices. This collaborative environment promotes innovation and drives business value.”
- Strategic partnerships: “Our COE manages and develops strategic, technical partnerships, enabling us to stay at the forefront of AI and data analytics trends.”
- Certification and training: “Our COE provides flexible and rigorous training, supporting project delivery and ensuring that our teams are equipped with the necessary skills to succeed.”
- Proposals and solutions: “Our COE supports proposals by providing technical solution strategies, enabling us to deliver innovative solutions to our customers.”
- Culture of innovation and creativity: “Our COE has been instrumental in fostering a culture of innovation and creativity, empowering employees to pursue their passions and drive business value.”
What Organizations Can Expect from an AI COE
The expectation of any COE is that it will help drive innovation and improve efficiency first and foremost, says the CTO of AI of a large data storage vendor. It is also important that a COE enhances the decision-making process and facilitates a healthy collaboration between business units and external partners.
At the storage vendor, “The COE is ultimately designed to move AI from theoretical exploration to practical implementation, delivering tangible business value by optimizing the many processes, enhancing efficiency, and unlocking data-driven insights,” the CTO of AI explains.
It should also be noted that by developing and implementing AI-driven solutions, the COE contributes to the creation of new revenue streams and business opportunities, he says.
An ideal COE should serve each business unit with both operational transparency and agile governance, Masood explains. At UST, the company places data engineers side by side with financial analysts to ensure swift translation from concept to execution excellence.
“Boards want tangible ROI — like the significant improvement in operational efficiency we saw once advanced ML was integrated into supply chain optimization,” Masood explains. “By embracing this purposeful approach, the COE stands at the center of a broader innovation ecosystem.”
Gaining Resilience and Competitive Edge
Any organization aiming for market and competitive resilience should consider developing an AI COE, Masood says.
“I’ve seen it become a vital change agent that champions iterative prototyping, fosters collaborative innovation, and sustains a high-performance culture,” Masood explains. “Companies with strong AI governance can boost [significant] returns on invested capital. C-suites increasingly view these [gains] as the hallmark of mission alignment, especially when shareholders demand better liquidity analysis and more reliable revenue management.”
An AI COE helps accelerate AI adoption by providing a dedicated hub for a more practical application, the storage vendor CTO of AI says. It achieves this by piloting innovative AI solutions tailored to specific industry needs, as evidenced by successful advanced prototypes in energy, industrial, and mobility industries, among others. This hands-on approach enables clients to see impactful results and understand the real-world implications of AI.
“The COE fosters collaboration with key technology partners, ensuring access to cutting-edge AI capabilities and facilitating the development of robust and impactful AI solutions that directly address client business challenges and unlock new value streams,” the CTO of AI explains. “This collaborative approach allows us to develop and deliver impactful prototypes, as we’ve already demonstrated in the energy, industrial, and mobility sectors, ultimately accelerating AI adoption and delivering tangible business value for our clients.”
Defining and Measuring Success
To determine if an organization has benefited from establishing an AI COE, it must first define what it means by success, Torzynski says. He recommends these steps:
- Establish clear goals and objectives: Define the purpose of the AI COE, such as improving operational efficiency, enhancing customer experience, or driving revenue growth.
- Develop key performance indicators (KPIs): Measure gains from the AI COE, such as with project completion rates, customer satisfaction, or revenue growth.
- Establish a framework for success: Outline the key elements that enhance the business, such as with innovation, impact, and efficiency.
To measure actual success, Torzynski says organizations should:
- Track project completion rates: Monitor the progress of each project and measure the impact on business outcomes.
- Monitor customer satisfaction: If customers of the COE have been identified, evaluate customer satisfaction and measure the impact of AI on customer experience.
- Measure revenue growth: Determine what portion of revenue growth can be attributed to the AI COE.
- Track innovation and impact: Assess the innovation and impact that AI projects have had and measure the value created.
- Conduct regular evaluations: Routinely assess the performance of team members and identify areas for improvement.
Finally, Masood stresses that a robust AI COE is more than just a discrete function; it becomes a cornerstone of enterprise architecture, shaping tomorrow’s go-to-market strategy and fueling future-proofing.
“The ability to harness frontier AI, such as generative models or agentic automation, hinges on forging an environment of continuous learning,” Masood explains. “The strategic imperative is clear. Organizations should establish their AI COE now to lead, rather than follow.”
(This is the first of a two-part series on AI COEs. In part 2, we explore how an AI COE impacts individual units in an organization, who should be selected to the COE team, and who the ideal leaders for the effort are.)
