Domain 4 review: readiness and transformation

Domain 4, Business Readiness, Leadership, and AI Transformation, carries 24 of every 100 scored points and has the most skills of any domain: four tasks and nineteen skills. Andre found it the most familiar in spirit, because operations managers spend their lives on readiness: whether the staff, systems and procedures are in place before something new starts. This lesson reviews the first two tasks, assessing readiness and maturity and establishing data and infrastructure foundations. Lesson 8 covers leading change, building workforce capability and scaling.

Task 4.1: readiness across dimensions

Skill 4.1.1 asks you to assess readiness across dimensions the guide names as examples: leadership alignment, data quality, culture, technical infrastructure and governance. A readiness assessment is the honest answer to a simple question: if we approved this AI programme tomorrow, what would stop it succeeding?

In the autumn, Andre and Noor had scored Cascade Ridge on these dimensions. Leadership alignment was strong after the board's request; data quality was mixed, because appointment records were good but pharmacy product codes differed between three systems; culture was cautious, especially among technicians who had heard rumours that AI would replace them; technical infrastructure was improving with the move to AWS; and governance was new, with an AI policy drafted but not yet approved. The exam often gives a scenario like this and asks which gap to address first. The answer is usually the gap that blocks the specific initiative being discussed, not the one that sounds most impressive to fix.

Maturity models

Skill 4.1.2 asks you to use AI maturity models to place an organization on its AI journey, from experimentation to enterprise-scale deployment. Maturity models differ in their names and number of stages, but they share a common shape.

The exam point in skills 4.1.3 and 4.1.4 is that investments should match maturity. An organization at the exploring stage gains little from an enterprise-wide agent platform; it needs literacy, policy and a few well-chosen pilots. An organization already running several systems needs monitoring and operating discipline more than another pilot. A good answer identifies capability gaps in people, process, technology and governance that block progress, then prioritizes investments that close the gaps on the path to the organization's strategic goals.

Task 4.2: data foundations

Skill 4.2.1 asks you to evaluate data readiness, including quality, accessibility and the effect of data silos. Data readiness is always judged for a specific use: data that is perfectly good for monthly reporting may be useless for a daily prediction because it arrives too late. The dimensions Andre used were completeness, accuracy, timeliness, accessibility, ownership and lawful use.

Data silos, where data sits in separate systems owned by different teams with different definitions, are among the most common blockers in exam scenarios. Cascade Ridge's three pharmacy systems used different product codes for the same medicines, so a demand forecast could not combine their history until someone built and maintained a mapping between them. The business fix is as important as the technical one: an owner for the master product list and a rule that new products get one code everywhere.

Skill 4.2.2 covers data strategy, data ownership and data sharing frameworks. A data strategy states which data matters most to the organization's goals and how it will be managed. Ownership assigns a named business owner, often called a data owner or data steward, who is accountable for each important dataset's quality and use. Sharing frameworks set the rules for using data across teams or with partners: purposes allowed, consent and privacy conditions, security, and who may approve new uses. Proprietary data governed this way is also a source of competitive advantage, which links back to Domain 2.

Infrastructure foundations

Skill 4.2.3 asks you to assess the foundational technology and infrastructure AI needs. At a business level, that means a secure environment where data can be stored and processed, ways to connect AI services to existing systems, the ability to monitor and control use and cost, and enough capacity for expected demand. Cloud platforms reduce the upfront investment: instead of buying hardware, an organization can use managed AI services and pay for use.

Remember the out-of-scope list from lesson 1: you will not be asked to configure any of these services. You may be asked which kind of capability is missing, or which service category fits a business need.

Practice scenarios for readiness and foundations

Readiness questions usually describe an organization in a few sentences and ask what to assess, what is missing or what to invest in next. The skill is to read the description as a readiness profile.

Data sharing frameworks in practice

Exam scenarios sometimes involve sharing data with a partner, such as a supplier, a research institution or another company in a joint venture. A sound data sharing framework answers a short list of questions before any data moves: what the data will be used for and what uses are forbidden; what legal basis or consent covers the sharing; what form the data takes, such as aggregated, de-identified or full records; how it will be secured in transit and at rest; who may approve new uses; how long it will be kept; and what happens to it when the arrangement ends. Andre found it helpful to picture Cascade Ridge sharing appointment trends with a pet food supplier. Aggregated, de-identified trends by region might be acceptable under a written agreement. Individual client records would not be.

Data ownership questions follow a similar pattern. The expected answer usually assigns a named business owner for each important dataset, because quality problems persist when everyone uses the data and no one is accountable for it. Technical teams maintain the systems; business owners decide definitions, quality standards and who may use the data for what.

Finally, remember that readiness is not a one-time verdict. An organization can be ready for a low-risk document extraction pilot and unready for a customer-facing agent at the same moment. Good answers judge readiness for the specific initiative in the question.