EmbedHow does proven value become a working model?
Turns the first application's experience into the organization's own way of working
Through the Proof stage, the organization has seen that AI can create value with real users on a selected business problem. The Embed stage combines that experience with what EnterpriseAIX brings from other organizations and carries it into a way of working the organization owns, one that more people, workflows and decision mechanisms can use.
Here the organization does more than use an application. Together, we set up how it will think, work, decide and build new applications with AI — and the place where those decisions are made: the Enterprise AI Transformation Office, or EAT Office. This is the lasting gain of the Embed stage; roles, principles and measurement find a home here. In our presentations this working model is called the AI Harness: the organization's first AI Harness is built in the Embed stage.
The Proof stage shows the first value.The Embed stage turns that value into a structure the organization can reproduce with its own people, decisions and workflows.
The structure that is built: Enterprise AI Transformation Office
A lightweight governance structure that brings transformation decisions together across leadership, business units and technology and governance teams. It is not a new department; it is clarity on decisions, responsibilities and follow-up cadence. Its first working principles are set in the Embed stage:
How the Embed stage runs
Over eight weeks, the experience handed over from the Proof stage is combined with EnterpriseAIX's accumulated practice, the EAT Office is established and the method is tested on a second application.
- 01
Week 1 · Consolidating what was learned
The experience handed over from the Proof stage — usage habits, trust conditions, value findings, roles — is put on the table with what EnterpriseAIX brings from other organizations. The raw material of the way of working comes from here.
- 02
Week 2 · First draft of the way of working
Roles, the division of work between people and AI, usage principles and the decision model are designed in draft.
- 03
Weeks 3–4 · Establishing the EAT Office and choosing the second application
An executive representative, business owners, Business Technologist candidates and technology and governance representatives are named; decision rights, agenda and follow-up cadence are set. The EAT Office's first decision is the choice of the second application opportunity; build starts the same week.
- 04
Weeks 5–6 · The second application in the EAT Office cadence
The build continues with the Forward Deployed Engineer approach; the EAT Office runs its first follow-up cycle on this application. The draft is tested and corrected on real work; role-based AI literacy begins.
The Forward Deployed Engineer role → - 05
Week 7 · Validation
Usage, value, trust and human oversight findings from the second application are assessed in the EAT Office; the model's fit with real working conditions becomes visible.
- 06
Week 8 · Operating model and next decision
Governance, value tracking and opportunity management become standing items on the EAT Office agenda; the decision on the Scale stage is made here.
How visible is AI literacy at this stage?
In the Proof stage, selected users learned to work with AI through real work. The Embed stage turns that experience into a development approach that can spread and be reused across the organization.
AI literacy is addressed in three layers: a shared AI language for the employees and managers of the relevant business units; for the application's users and process owners, the skill of framing problems, setting context and assessing output in their own workflows; for Business Technologist candidates, the ability to define opportunities, bridge the business owner and the solution team, and track business value.
Outputs of the Embed stage
At the end of the Embed stage, the organization keeps:
- EAT Office — a governance structure of executive leadership, business owners, Business Technologists and technology and governance representatives that has made its first decision and run its first follow-up cycle.
- The organization's own working model — a repeatable AI working practice tested on two applications.
- Role and responsibility framework — a first capability definition for the Business Technologist profile and a model of business ownership.
- Division of work between people and AI — usage principles, approval points and a human oversight model.
- Forward Deployed Engineer way of working — how business and technical teams work side by side.
- Governance principles — workable rules for trust, risk, access and compliance.
- Value measurement and improvement cadence — regular review of usage, business value and user feedback.
- Opportunity register — the structure in which new opportunities are recorded on a standard opportunity card, and which becomes the Portfolio.
What happens if the Embed stage succeeds?
Success is not that the organization has a second AI application; it is that the following structures actually begin to operate:
- Principles for how people and AI work together that hold across at least two applications.
- Clear business ownership, user responsibility and decision authority.
- People inside the organization named for the Business Technologist role and at work.
- An EAT Office that has made the second-application decision itself and run at least one follow-up cycle.
- Regular review of usage, business value, quality and risk indicators.
- New opportunities recorded and prioritized to the same standard.
- AI literacy that has become role-based development rather than one-off training.
What the Embed stage does not do
- No new department is created; the EAT Office starts with responsibilities taken on by existing staff.
- Not everyone receives the same generic AI training; AI literacy is role-based.
- The second application is not chosen as a copy of the first.
- Software licenses, model usage fees and infrastructure investment are not included.
- Broad system integration and additional development the organization may request are not carried out.
Commercial terms of the Embed stage
The standard investment covers structuring the learnings from the first application, the Business Technologist profile and role framework, role-based AI literacy tied to the application, the EAT Office's first working and decision cadence, the division of work between people and AI, usage principles and the trust framework, value measurement and continuous improvement, the treatment of the second application for this stage, and the closing session.
For multi-company groups, heavy compliance requirements, a large user base or a technically more complex second application, the duration extends to 10–12 weeks.
Scale
The Embed stage establishes the organization's way of creating value with AI and the EAT Office. In the Scale stage, responsibility passes to the organization's teams; with the tailored method, value is uncovered in each department according to its own processes and needs.