<img height="1" width="1" style="display:none" src="https://www.facebook.com/tr?id=266259327823226&amp;ev=PageView&amp;noscript=1"> Skip to content

AI Center of Excellence (AI CoE)

AI Center of Excellence (AI CoE) is an internal support function. It ensures that AI is applied where it creates real value for the organization, and that AI work has a shared direction so that projects and knowledge reach the whole organization rather than remaining in silos.

Norrin builds AI CoEs starting from the organization's goals, scaling the governance model from lightweight to more structured as maturity grows.

How should an organization implement an AI CoE?

  • Start with a lightweight governance model

    An AI CoE does not require a large investment from the start. You can begin with a lightweight governance model and grow it as maturity increases. Even a simple structure is enough when responsibilities are clear and metrics are in place. Organizations that build support functions like this consistently outperform those where AI projects remain isolated.

  • AI CoE

    Set roles according to the organization's needs

    The structure and roles of an AI CoE are shaped by the organization's needs, culture, and goals. Before setting it up, the current state is assessed: AI strategy and leadership commitment, organizational readiness for change and staff capability, ongoing AI projects, and data quality, availability, and the condition of existing systems.

  • AI CoE prioritizes projects based on leadership direction

    Leadership sets the AI strategy and goals. The AI CoE ensures that day-to-day work is consistent, supports those goals, and that choices made by different teams are not working against each other. The leadership team does not need to spend time on individual decisions since those are the AI CoE's responsibility.

  • Manage risk and compliance centrally

    All AI projects go through the AI CoE, which evaluates every solution before implementation. The AI CoE ensures compliance is in order and that obligations such as the EU AI Act, GDPR, and others are accounted for.

  • AI CoE supports leadership, managers, and employees in their daily work

    The AI CoE reports regularly to leadership on progress and challenges, gives managers information on new solutions and support for change management, and gives employees a channel to bring their ideas forward. It is the place everyone knows to turn to on AI matters.

What organizations should know before setting up an AI CoE?

  • Assess the current state before establishing an AI CoE

    Before setting up the structure, you need a clear picture of what you want to achieve, where you are today, and where the biggest development needs are. This can be done through interviews, workshops, or surveys, and the AI CoE is then built on that foundation. Without this assessment, the AI CoE risks becoming a disconnected structure without a clear purpose.

  • Build the AI CoE as a support function

    An AI CoE is a support function that serves the entire organization, just as HR handles people matters for all teams. Unlike a project, it is a permanent structure that carries responsibility for AI work across the whole organization.

  • AI CoE

    Set goals for AI, don't follow the trend

    In many organizations, AI work happens because others are doing it, not because the best use cases have been identified or goals defined. Without clear goals, AI work fragments into isolated experiments that generate neither organizational learning nor measurable benefit.

  • Lead change as part of the AI CoE's work

    Change management is the AI CoE's core task. Without it, AI does not take hold in the organization. People are left uncertain about their role in AI projects, and concerns about job security can grow. The AI CoE creates transparency across the organization and ensures that new solutions are actually adopted.

How we build an AI CoE?

Initial assessment

Before building the AI CoE, we need a clear picture of where things stand. We map the current state together through interviews, surveys, or workshops. This tells us what conditions will support the AI CoE's success, where it should focus, and what the biggest obstacles and development areas are.

Synthesis

From the assessment, we build a shared picture of what the initial mapping revealed, which problems need to be solved first, and how. The synthesis may show, for example, that the organization lacks a shared way of working or a clear vision for AI's role. We resolve those before the structure is put in place.

Building the AI CoE

Once the current state is clear, we build the AI CoE together with the client. The goal is a clear structure where roles and responsibilities, processes, metrics, and change management are designed to fit the organization.

Could we help you with the AI CoE?

Or leave us a message!