AI Native Transformation Framework
A cycle with decision points.
Each stage removes a different uncertainty: where you are, where you want to be, whether it’s viable, whether it delivers value in use, and whether it can scale. Risk controls, adoption and measurement run through the whole cycle.
- 01
Assess
Main question
Which problem is worth solving
Exit decision
Agree on problem and evidence
- 02
Envision
Main question
How the business should operate
Exit decision
Choose capability and business case
- 03
Architect
Main question
How to make it viable and controllable
Exit decision
Authorize design and investment
- 04
Activate
Main question
Does it work with real users and data
Exit decision
Go, Iterate or Stop
- 05
Scale
Main question
Can it be sustained and extended
Exit decision
Expand under controls
Working rules
- Start from a business outcome and a bounded process.
- Write down hypotheses and assumptions before choosing tools.
- Set a baseline and a named owner for every indicator.
- Prefer the least complexity that solves the problem.
- No approval is implied: the sponsor can stop the project over an unaccepted risk.
Go, Iterate or Stop
In Activate, the pilot is judged against thresholds agreed before it starts. A critical security failure blocks Go even if average quality clears the bar.
The six dimensions of the index
- D1
Business & strategy
Objectives, sponsorship, investment and value
- D2
People & culture
Skills, adoption, leadership and roles
- D3
Processes & operations
Flow, ownership, exceptions and measurement
- D4
Technology & architecture
Integration, operations and technical decisions
- D5
Data & knowledge
Quality, access, ownership and lifecycle
- D6
AI readiness & governance
Inventory, evaluation, control and oversight