Not sure your AI can be trusted? Take the 2-minute Reliability Quiz. Start the quiz
Phase 2 · Fix the foundation

The Knowledge Foundation Build

The work that repairs the foundation, so your content stays reliable for people and AI alike.

Book a 30-minute call

Structure before content.

Once the audit shows what is broken, this phase builds the fix. The principle is simple and it is not negotiable: structure before content. I define the structure, the ownership, and the lifecycle first, then content flows into a system that keeps it accurate and current by design. Writing more documents into a broken system only makes a larger mess.

One thing this build asks of you: a named person inside your organization who will own the system once it is running. A framework with no internal owner becomes a document everyone admires and no one operates. So naming that owner is part of the work, and the final phase coaches them to run it confidently.

The sequence

Where this sits.

Phase 1
AI Reliability Audit

Diagnose. Trace the failures to their root and set the order of operations.

Phase 2
Knowledge Foundation Build

Fix the foundation: structure, ownership, and lifecycle.

Phase 3
AI Enablement Program

Bring people to the tools, with adoption measured.

When this is the right place to start.

Your content is fragmented, contradictory, or unowned.

There is no single source of truth, so people cannot tell which version to trust.

Knowledge does not survive when people leave.

A key person is leaving or retiring, and what they know is not written down anywhere.

An audit has just shown you the gaps, and now they need closing.

Your knowledge is well kept for people, but not structured so an AI can retrieve it reliably.

The build

Three pillars, built in sequence, then activated.

The order is the method. Each pillar depends on the one before it.

01
Structure.

The information architecture beneath your content: how it is organized, the types of content you have, how it is labeled so it can be found, the naming conventions, and a pilot that shows the model working in your platform. This is the skeleton. People need a system to work in before they can be accountable for it.

02
People.

An ownership model, defined roles, and review cadences, with one named person accountable for the system overall. Content without a named owner and a review date decays. Ownership is distributed rather than centralized, so the system scales beyond any single person.

03
Process.

The content lifecycle, contribution workflows, and quality checkpoints. This is what keeps the foundation alive after I leave.

A note on design Every element is built to serve two audiences at once, the people navigating your content and the AI that queries it. That means the foundation is ready for AI retrieval without a separate technical layer bolted on later.

The deliverables

What you receive.

01

A knowledge framework: the authoritative rules, from metadata to lifecycle to naming.

02

A roles playbook: what each role does, when, and how, with checklists and templates.

03

A definition of the owner role, which doubles as the specification if you decide to create the position.

04

A working pilot set up in your platform (Confluence, Notion, SharePoint, or ServiceNow) that shows the model in action, validated with you.

05

A content lifecycle model with clear stages and rules.

06

A configuration guide and a 90-day implementation plan, sequenced to your existing workflow, that your team follows to roll it out with my support.

07

A 90-day Activation phase: fixed-cadence coaching of your named owner, with a few simple measures tracked and reported, so the system is actually running before I step back.

This is design, a pilot, and activation support. It does not include migrating your existing content or reconfiguring your platform end to end. Your content owners move content into the new structure over the 90 days, guided by the framework and coached through it.

What changes

A single source of truth, reliable by design.

You get content your AI can draw on dependably, and ownership that survives turnover. The knowledge base stops being a place things go to be lost, and someone inside your organization can run it without me.

Build the foundation your AI depends on.

The first conversation takes 30 minutes. It is diagnostic, not a pitch: a chance to understand your content, your platforms, and what the build would involve.

Book a 30-minute call