Audit made the four layers of your people’s work visible. Activate shows where those layers actually sit when decisions get made, so you can put AI where it counts most.
Activation means building your systems out to hold knowledge they’re missing today.
Close the gap, and the project moves right.
A tool that reviews contracts the way a senior lawyer does. It flags the unusual clause, applies the precedent, calls the risk. Then it stalls.
The reason is simple. The judgement it copies is in one person’s head, and the AI has nothing to work from.
Ready to work. But it has nothing to work from.
The senior lawyer’s call. In one person’s head.
How much does it matter? And can the AI use the knowledge it needs? The answers put every project in one of four boxes. Not every project should aim for the top.
High-stakes calls, but the judgement is locked in someone’s head. The answers look right and miss what decides the call.
Your documented knowledge and your experts’ judgement, both given to the AI. It hands back a view a person can question. Where the return builds and lasts.
Standard, off-the-shelf help: summarising, drafting, generic templates. Easy time savings, but not where you build an edge.
AI working straight from your documented data, pulled in as it answers. A reliable base for the box above.
Not every project should aim for the top. A “where’s my order?” lookup belongs in Foundation. It is low-stakes and the AI can already use the data. Dragging it toward Advantage just wastes effort. The contract tool is the opposite: it matters a lot, and today it sits stuck in Context Debt.
To move a project from Context Debt to Advantage, the AI has to draw on each layer of your knowledge in turn. The first two it can use today. The last one is where the work is.
What any AI does out of the box. General craft, already in the model.
Written down and connected, so the AI can use it directly.
Still in people’s heads. The AI has nothing to work from. This is the gap.
You don’t close it with a better model. You codify it. That’s your moat.
The model gets better on its own, every release, whether you do anything or not. Your codified knowledge is the part that is yours, and it is where the real return is built.
Getting people to trust and act on what the AI produces is its own discipline. That is the Verification pillar.
Two interactive walkthroughs: a contract-review project, and a save-or-let-go customer decision, taken through the map step by step. Leave your email and both open straight away.
One email. The two walkthroughs, and the occasional note when something new lands.
Each project gets a place. Re-plot them as the work moves and the knowledge gets codified. Check it quarterly.