Skip to main content

The Lexicon

The meanings we work to.

A standard is only as strong as its vocabulary. These are the terms AIRL™ is built on: the framework's own, defined here canonically, and the practice terms they inherit from the Kaipability lexicon.

Framework Terms

AIRL™

The AI Readiness Level framework: nine levels across four stages measuring an organisation's capability to deploy AI at economic scale. Trademarked so the definitions cannot drift; a level only functions as a standard if everyone means the same thing by it.

Readiness Level

A position on the nine-level climb from Feasibility Identified (1) to Operational at Economic Scale (9), passed through evidence gates, never self-declared.

Gate

The evidence test between one level and the next. Measured QCDS improvement, security audits, validated economics. A gate is passed or it is not; there is no partial credit.

QCDS

Quality, cost, delivery, safety: the four metrics every AIRL level is gated against, and the language operations already runs on.

The Four Stages

Concept (levels 1–3), Development (4–5), Demonstration (6–7), Production (8–9). The same arc every deployed capability travels, from feasibility to autonomous operation at economic scale.

Indicative Level

The level suggested by a self-assessment. Useful as a starting position, never as a certificate: indicative levels are capped at 8 because Level 9 requires twelve months of sustained, evidenced operation.

Gated Assessment

The real thing: the same three dimensions as the self-assessment, with self-reporting replaced by evidence, delivered on site by Manufacturing Engineers.

DaaS — Deployment as a Service

Engagements scoped as movement through gate criteria and priced on outcomes the customer can audit, not on seats or licences. The structural break with SaaS.

Adoption vs Deployment

Adoption is a purchasing decision. Deployment is an engineering achievement. Most of what industry calls an AI deployment is adoption: AIRL exists to measure the difference.

Manufacturing Engineer

A practitioner of Manufacturing Engineering — "the discipline of making things makeable, turning a design into a system of people, machines and methods that can produce it reliably, at viable cost." The people who deliver AIRL assessments.

Practice Terms · Shared with kaipability.com

Manufacturing Engineering

The discipline of making things makeable, turning a design into a system of people, machines and methods that can produce it reliably, at viable cost.

Modern Industrialist

An operator who treats industrial capability as the primary asset to build, own and compound, accountable to repeatable real-world outcome, not activity.

Physical AI

Artificial intelligence that senses, decides and acts on the physical world, machines, robots, processes, not only on data or text.

AI-native

A production system designed so the process and its AI assume each other from the outset, built in, not bolted on.

Deployment Readiness

The honest measure of how close a capability is to producing its intended outcome reliably, repeatedly, and at viable cost in real conditions, not in a lab, not on a good day.

Capability

The ability of a specific human-and-machine system to repeatedly produce a specific outcome, to a known standard, under realistic conditions. The asset that compounds, not the patent.

Valley of Death

In industry, a capability gap, not a funding gap. Where most hard-tech scale-up fails between a working technology and reliable production.

Manufacturing

The disciplined turning of materials, energy and information into useful goods, repeatedly, to standard, at viable cost.

Every one of these terms is doing work somewhere in the framework. The fastest way to see them in action is to take the assessment.

Take the Assessment