Core Concepts
AI Native Organisation
An organisation whose knowledge, workflows, decision rights, governance, products, services, and learning loops are designed around human-led machine intelligence.
Read more →
Knowledge Bank
A working library
for AI-native practice.
Definitions, concepts, and reference patterns we use across advisory engagements — written for leaders, not vendors.
In the library
37entries
Across 8 categories.
Showing 37 of 37
Core Concepts
Organisational Intelligence
The capacity to sense reality, remember what it knows, reason across functions, decide in time, act coherently, and learn from consequences.
Read more →
Core Concepts
Organisational Unintelligence
The condition of abundant data, dashboards, and expertise that nonetheless cannot become timely, reliable, accountable intelligence.
Read more →
Core Concepts
Queryability
The ability of authorised users to ask meaningful questions of the organisation and receive timely, permissioned, cited, explainable answers from trusted sources.
Read more →
Core Concepts
Human Middleware
People manually bridging gaps between systems, processes, teams, and knowledge stores.
Read more →
Core Concepts
Decision Rights
Clear rules about what AI may decide, recommend, draft, execute, escalate, or never touch.
Read more →
Core Concepts
Knowledge Operations
The discipline of maintaining organisational knowledge as a living operational asset: source quality, ownership, freshness, provenance, retrieval, permissions, feedback.
Read more →
Core Concepts
Agentic Workflow
A workflow decomposed into steps where some are delegated to AI agents within explicit boundaries, with human handoffs at consequential moments.
Read more →
Operating Model
AI Strategy Layer
Where the organisation decides what AI is for: which outcomes, which constraints, which non-goals.
Read more →
Operating Model
Knowledge Layer
Where organisational memory becomes queryable, governed, and trusted enough for decisions.
Read more →
Operating Model
Workflow Layer
Where work is decomposed and AI is inserted with boundaries, handoffs, and evaluation.
Read more →
Operating Model
Governance Layer
Where decision rights, risk tiers, approvals, and incident response are defined and enforced.
Read more →
Operating Model
Capability Layer
Where roles, skills, and organisational learning are redesigned for AI-augmented work.
Read more →
Operating Model
Technology Layer
Models, vendors, integrations, deployment patterns, and the architecture that holds them.
Read more →
Governance and Risk
Non-Automation Register
A named list of decisions and tasks that must remain human, with reasons recorded.
Read more →
Governance and Risk
Algorithmic Impact Assessment
A structured assessment of who is affected by an AI system, what could go wrong, and how harm is mitigated and contested.
Read more →
Governance and Risk
Contestability
The ability of users and affected parties to challenge, correct, and appeal AI-assisted outputs.
Read more →
Governance and Risk
Audit Trails
Durable records of what was asked, retrieved, recommended, decided, and overridden.
Read more →
Governance and Risk
Source Provenance
Knowing where each piece of information used by an AI system came from, and when.
Read more →
Governance and Risk
Evaluation Harness
A repeatable test suite of representative inputs, expected behaviours, and failure modes.
Read more →
Workflow Redesign
Task Decomposition
Breaking a workflow into discrete steps with explicit inputs, outputs, and decision points.
Read more →
Workflow Redesign
Human Handoff
A deliberately designed moment where work moves from AI to human (or vice versa) with full context.
Read more →
Workflow Redesign
Exception Handling
Explicit rules for what happens when the AI is uncertain, the input is unusual, or the output is contested.
Read more →
Workflow Redesign
Outcome Metrics
Metrics that measure the change in actual outcomes — not just throughput or activity.
Read more →
Workflow Redesign
Agent Boundaries
Explicit, written limits on what an AI agent may access, do, and decide.
Read more →
Knowledge and Data
Knowledge Map
A view of where institutional knowledge lives, who owns it, how fresh it is, and how it is used.
Read more →
Knowledge and Data
Source Quality Rubric
A shared standard for what makes a knowledge source good enough to be cited by an AI system.
Read more →
Knowledge and Data
Stewardship Model
Named owners responsible for the freshness, accuracy, and lifecycle of each knowledge domain.
Read more →
Workforce and Culture
Capability Ladder
A progression from AI-literate, to AI-fluent, to AI-leading roles across the organisation.
Read more →
Workforce and Culture
Role Redesign
The deliberate redesign of a role’s purpose, tasks, decision rights, and metrics in light of AI.
Read more →
Procurement and Vendors
Build-Buy-Partner Decisions
A structured choice between building internally, buying off-the-shelf, or partnering with specialists.
Read more →
Procurement and Vendors
Vendor Lock-In
Dependence on a vendor that is costly or impractical to exit.
Read more →
Procurement and Vendors
Model Portability
The ability to move from one model or provider to another without rewriting the system.
Read more →
Procurement and Vendors
Contractual Safeguards
Terms covering data use, model training, security, audit rights, exit, and incident response.
Read more →
Procurement and Vendors
Public-Interest Procurement
Procurement designed to meet public-interest standards on transparency, fairness, contestability, and value.
Read more →
Case Studies
Estonia X-Road
Interoperability and digital identity infrastructure that lets services compose coherently.
Read more →
Case Studies
Australia Robodebt
Automated welfare debt inference that bypassed lawful, fair, human-centred process.
Read more →