Personal productivity
Drafting and rewriting. Useful, but the thinking was still entirely mine.
AI LAB
This is not a software portfolio or an engineering showcase. It is an environment for capability development and evidence-based innovation — small, working systems built to test whether the methodology in the Business Library still holds when a machine executes it.
Everything here gets published as it is built — including the parts that don't work.
My own use of AI moved through four phases. Each one increased the complexity of what AI was trusted to help build — not just what it was asked to produce.
Drafting and rewriting. Useful, but the thinking was still entirely mine.
Contract and document review, where the value was in what I might have missed.
Structuring fifteen years of engagements into frameworks that hold up under executive scrutiny.
Building working systems — where the constraint stops being execution and becomes thinking.
Three projects, each built against a problem I have actually managed — not a demo. Each will ship with the code, an honest write-up of what worked, and what it still gets wrong.
The problem: health scores tell you an account is declining after it has already declined, and nobody trusts the number because it can't show its evidence.
What it does: takes raw customer notes and produces a governed diagnostic result, evaluated by a deterministic rules engine against confirmed evidence, then explains that result in plain language. The AI interprets and communicates; it never decides the outcome.
The problem: most QBRs report activity to a customer who wanted evidence of outcomes. Preparing one properly takes hours that portfolio managers do not have, so the quality degrades exactly where the account is most at risk.
What it does: turns account data into an executive-ready review structured the way I ran them — business objective, what changed, evidence, decisions required — in the customer's language rather than the vendor's.
The problem: most teams can describe what their process does but not why it produces the result it produces. Diagnosis is the expensive part, and it usually requires someone who has seen the pattern before.
What it does: takes a described process and returns the likely constraint, the questions worth asking next, and a prioritized intervention list — applying the Operational Excellence Cycle instead of generic best practice.
I could have waited and published this page only once it was full. But the first principle in the Business Library is that there is always a better way — and the second is to use evidence, not opinions. Naming what I am building, with a date, is the evidence. The first project going live is the proof; the same standard holds for what is still in build.
VIEW THE CHDM REPO ON GITHUBThese projects are not experiments in what AI can do. They are tests of whether fifteen years of operating methodology can be encoded well enough that a machine applies it usefully. The methodology itself lives in the Business Library.
The Latent Value Model, the Technology Leverage Ladder, and three engagements where technology was never the constraint.
READ THE METHOD ALL FOURGovernance & Risk, Operational Excellence, Customer Success, and AI & Intelligent Operations.
EXPLORE REAL TRANSFORMATIONS