Start here Operating model
What is an AI-native business?
Working definitions for AI-native, AI-first and AI-enabled, with practical traits and an illustrative operations example.
Read the guideExplore the guides
03 more perspectivesGovernance
Risk-classifying AI agent actions
An example action-risk policy, with review responsibilities, data-access considerations and implementation checks.
Read the guideOperations
Decision latency in multi-site operations
A four-timestamp method for measuring operational delays, with a clearly labelled hypothetical restaurant example.
Read the guideBuyer's checklist
Questions to ask AI agent vendors
Twenty questions covering data, permissions, approvals, models, operation and handover, with evidence to request.
Read the guidePut the ideas to work.
These articles explain terms, provide worked examples and suggest questions for a first implementation. The focus is the business task: what information it needs, where review belongs and how to judge whether the result is useful. Use the resources below to turn that understanding into a practical next step. Choose one recurring task with clear inputs and a person responsible for reviewing the result.
- Prompt pack & playbooks — Choose a workflow, prepare useful prompts and set measurement criteria before you begin.
- The AI glossary — Clear definitions and everyday examples for the technical terms you will encounter.
- How implementation works — See how the scope, inputs and review criteria come together around a real business task.
Examples are identified as illustrative, and proposed policies need to be adapted to your organisation. A general guide cannot establish that a particular product configuration or customer workflow has the controls described.