People × Processes × Data × AI: Assessing Readiness for Sustainable Transformation

Technology on its own does not transform organizations. People do. After more than 15 years across software engineering, architecture and technology leadership, I am turning the pattern I trust most into an assessment model: People × Processes × Data × AI = Readiness for AI-enabled transformation.

Its immediate output is readiness for a defined initiative, not a promise of results. It helps structure the assessment and identify essential gaps before implementation. Organizational outcomes are evaluated separately, through agreed functional quality criteria, expected and realized return on investment, and relevant sustainability impacts. The model is in development: I use its structure to frame assessments, while the evidence, scoring rules and thresholds still require evaluation before it can be treated as a validated instrument.

Why people come first

People establish the purpose, assume accountability and direct the processes. Processes use and generate data. AI uses relevant data to support the purpose the organization intends. The chain is not closed, so the assessment has to accommodate external data, pretrained AI and feedback between factors: an AI system that changes how a process runs also changes the data that process produces.

What the multiplication means

Each factor receives a proposed normalized score between zero and one. Their product is a readiness index, and multiplying it by 100 expresses it in index points. Two properties follow.

• A zero factor yields a zero product. If no one owns the outcome, or the data the initiative needs does not exist, it is not ready, however strong the rest may be.
• Above zero, higher scores can partially offset lower ones mathematically. A capable team and good data can lift the index while another factor is still modest.

The second property is why the index cannot decide on its own. Essential prerequisites, such as a named owner or lawful access to the data, have explicit decision rules, and a high overall index cannot override a failed essential prerequisite. Which evidence supports each score, how scores are assigned and where the thresholds sit all still require evaluation.

How I apply it in practice

• People: establish one accountable owner with authority over the intended outcome. Confirm role-specific competence and AI literacy, and define human oversight, including when people review and when they intervene.
• Processes: define the workflow, responsibilities, controls, exceptions and measures, so improvement can be observed rather than asserted.
• Data: assess suitability, quality, access and governance for the intended use, including external sources where relevant.
• AI: assess whether the task suits AI at all, then its integration, evaluation and operational requirements.

The owner approves the evaluation approach: the functional acceptance criteria, the ROI estimate and the later review of actual outcomes against both. Readiness says the initiative can start well. Those agreed measures say whether it was worth doing.

Sustainability runs through every factor

Sustainability is not a fifth factor. It shapes all four, across the economic, social, environmental and cultural considerations relevant to the initiative: a workload people can sustain, accessibility for everyone who uses the result, processes that hold without heroics, resource use proportionate to the value, ongoing costs known before adoption, and the effects on the people the change touches, including how they work together. A result that lasts one quarter was never a result.

An illustration

In the modernization of a regulated SaaS product already in operation for medical-device manufacturers, the gains did not come from a single technology. They came from coordinated organizational improvements: clearer accountability, an explicit delivery cadence and automation applied where the process was already understood. The delivery cycle went from four weeks to one, approximately 80% of manual customer-onboarding work was automated, and cloud infrastructure costs fell by approximately 80%, while the engineering organization grew by 50% and delivery stayed predictable.

I share it as an illustration of what coordinated improvements can achieve, not as evidence for the formula. The figures do not isolate the effect of each factor, and they do not validate the model.

Where to start

Start with a defined initiative and its intended outcome. Gather evidence for each factor rather than opinions about it. Identify the essential conditions that are not yet met, because they come first. Then prioritize the remaining improvements by risk, dependencies, feasibility and the outcome you intend. The lowest score is a signal, not automatically the roadmap.