Find the starting point
Stakeholder interviews, workflow evidence, and value hypotheses. Understand where AI can move performance first — before any tool is chosen.
The Client Transformation Methodology
Seven gates to an operating system.
AI fails when it is applied to a company it cannot read. The CTM turns adoption into a sequenced operating journey — not a one-off project. Each gate must be passed before the next opens.
Phase 1 · Diagnose
Find where AI moves performance first. Map the work and the context. Decide what stays human.
Map the work
Break workflows into discrete, measurable components. Map dependencies, bottlenecks, decisions, data, and value levers.
Classify human vs agent
Determine what stays human, what becomes agentic, and what needs a hybrid pattern. Not everything should be automated — this gate is where judgement is designed in.
Phase 2 · Design
Design the target operating system. Specify the controls and metrics that make it measurable.
Design the operating system
Target-state workflows, agent roles, decision rights, validation frameworks, and integration architecture.
Specify controls and metrics
Detailed specifications for each agent, integration point, edge trigger, risk control, and success metric. No ambiguity survives this gate.
Phase 3 · Deploy
Build scoped capability into production, then expand it one workflow at a time.
Build the agentic workflow
Rapid development in 30-day sprints. Working prototype from week one; production-ready scoped capability by week four.
Deploy, adopt, measure
Production deployment with outcome tracking, team training, operating rhythm, and continuous optimisation — measured against your own baseline.
"Start with legibility. Finish with operating capability."
Where do you start?
Your entry gate depends on what you already know.
If AI matters but the starting point isn't obvious.
Two to three weeks to a board-ready roadmap, an investment case, and a named first workflow.
If a specific workflow is ready and owned.
A bounded workflow with an accountable owner can enter the Sprint gate directly — prototype in week one.
If you're in private equity with a live data room.
AI Due Diligence takes 2–4 weeks, depending on access to people and data, and its AI-readiness score becomes the starting point for Gate 01.
Begin
Walk the gates with us.
Book a discovery call and we'll tell you — honestly — which gate you're standing at.