ScaleHow does a tailored method create value in each department?
Spreads the proven way of working across the organization's priority business areas
The aim of the Scale stage is not to bring an AI application to every possible department at once. It is to use the method built for the organization to reach each department's own processes and needs, uncover the AI-based value there, and turn it into a managed organizational capability.
In the Proof stage, EnterpriseAIX builds and proves the application. In the Scale stage, responsibility passes to the organization's teams: they run the work, and EnterpriseAIX stays on the team in an advisory and oversight role. The aim is not more applications; it is for every individual in the relevant departments to adopt a way of thinking and working with AI.
The organization no longer starts from scratch: there is a shared language for defining opportunities, the Business Technologist profile has begun to surface opportunities in business units, the Enterprise AI Transformation Office (EAT Office) runs the priority and decision cadence, and usage and trust principles have been learned from real applications.
Scaling is not about building more AI applications.It is about the organization's ability to deliver the right applications, in the right order, with the same working discipline.
Five working areas of the Scale stage
How the Scale stage runs
The Scale stage runs in cycles of 12–16 weeks. In each cycle, two applications move forward in two different business areas; the organization's teams run the work and EnterpriseAIX makes sure the method is followed.
- 01
Portfolio and priority — the EAT Office decides
The cycle's priority business areas and its two applications are chosen in the organization's own EAT Office; the Business Technologist network grows. EnterpriseAIX gives its view; the decision is the organization's.
- 02
Rollout to departments — the organization's teams deliver
Business Technologists surface opportunities in business units; the organization's teams build the applications and own them. EnterpriseAIX, on the same team, makes sure the method is followed, using the Forward Deployed Engineer approach.
The Business Technologist role →The Forward Deployed Engineer role → - 03
Shared language — for everyone, by role
All employees of the relevant departments receive a shared AI language, role-based usage guides and a regular knowledge flow. What spreads is not an application but a way of seeing.
- 04
Review and cycle decision — in the EAT Office
Usage, business value, trust and learning results are reviewed at the organization's own table; the Portfolio is updated; the decision on a new cycle is made. EnterpriseAIX gives its view.
How visible is AI literacy at this stage?
In the Scale stage, AI literacy does not mean giving everyone the same generic training. It spreads according to each business area's applications, risks and business goals.
Four layers move together: a shared organizational language for all employees and managers; for users in the cycle's business units, the practice of using their own applications and assessing output; a Business Technologist network of selected business-unit representatives; and a technical and governance community of IT, data, information security, legal and compliance representatives.
Outputs of the Scale stage
At the end of each cycle, the organization keeps:
- AI Application Portfolio — a living leadership view with business value, ownership, risk and priority.
- Applications working in different business areas — solutions used by real users and measured for value.
- Department rollout plan — the roadmap for the next cycle.
- The EAT Office's follow-up model — the cadence for tracking decisions, investment, value and risk.
- Business Technologist network — an in-house structure that surfaces opportunities in business units and supports users.
- AI literacy rollout model — role-based guides and a library of good practice.
- Knowledge flow and community cadence — the model that keeps what is learned with the organization rather than with individuals.
- Trust and compliance framework — human oversight and compliance principles for the applications.
What happens if the Scale stage succeeds?
Success at this stage is measured not by the number of applications but by the capability the organization gains. The outcome sought is AI Harness Capacity: the organization's ability to move forward on each new opportunity without starting from scratch — choosing the right problem, designing the division of work between people and AI, building applications safely and measuring their value. The last four criteria below are those abilities; the first two show that ownership sits with the organization.
- In-house ownership — are cycles running under the responsibility of the organization's teams, with EnterpriseAIX in an advisory and oversight role?
- Individual use — do employees in the relevant departments use AI in their own work, on their own, and within the organization's principles?
- Choosing the right problem — are opportunities defined in the same frame and comparable?
- Designing the division of work between people and AI — in a new application, are roles and approval points set up as a matter of course?
- Building applications safely — are access, human oversight and compliance handled with the same discipline in every cycle?
- Measuring value — are applications' usage and business value tracked in one shared leadership view?
What the Scale stage does not do
- No separate EAT Office is set up in each department; every department is reached with one structure and one method.
- Applications are not carried into every department at once; rollout happens in controlled cycles.
- The number of applications is not treated as a success measure.
- Not everyone receives the same generic training; AI literacy spreads by business area.
- Software licenses, model usage fees and infrastructure investment are not included.
- Work requiring broad integration, custom development or outside expertise is not taken on within this scope.
Commercial terms of the Scale stage
A standard Scale cycle lasts 12–16 weeks and covers the EAT Office cadence, portfolio management, AI literacy rollout and the rollout of two AI applications.
Applications beyond the two included in the cycle are assessed separately. A standard additional application is one that focuses on a single clear business problem, has a limited user group, can work with existing information sources and needs no extensive custom development.
The Scale stage is not an end point.It is the cadence that keeps the organization's ability to work with AI current and carries it into new opportunities.
A New Scale Cycle
At the end of each cycle, the portfolio, application results, the reach of AI literacy, trust conditions and the EAT Office's working cadence are reviewed together. Responsibility sits with the organization's teams from day one; EnterpriseAIX stays on the team in an advisory and oversight role, and its role shrinks with every cycle.