Methodology

The Practical AI Capability Model

Six dimensions, each scored from 0 to 100. A person can be fluent and still unsafe, or careful and still unable to frame a problem — so the dimensions are scored separately and never averaged away.

Sample profile

Green: current score. Grey: organisation benchmark.

Capability levels

Awareness0–19
Beginner20–39
Practitioner40–59
Proficient60–74
Advanced75–89
Capability Leader90–100

Every result also reports change since the last assessment, a confidence indicator, completion percentage, and a benchmark against team and organisation averages.

0182

AI Foundations

Understands model behaviour, terminology, strengths and hard limits, and can tell an appropriate use case from an inappropriate one.

0268

Responsible Practice

Handles confidential data correctly, recognises bias and security exposure, respects ownership, and keeps a human accountable for every output.

0361

Problem Framing

Converts a loose business ask into a clear problem statement with outcomes, constraints, stakeholders and success criteria.

0474

Workflow Design

Specifies inputs, instructions, approval gates, escalation paths, tooling and expected outputs so the work runs the same way twice.

0579

Applied Execution

Uses AI tools to complete practical tasks faster and better, and to support decisions inside real operating work.

0671

Quality Judgement

Checks accuracy, completeness, relevance, safety and bias, and knows when an output should not be used at all.