Skip to main content

Build Your Own KnowledgeFund

KnowledgeFund should not only exist as something a consultant does to an organisation.

It should also be something an organisation can learn to do for itself.

This page reframes the KnowledgeFund method as a self-implementation guide for internal leaders, operators, architects, and transformation teams.

Table of Contents​

The aim​

The aim is to help an organisation build the internal capability to:

  • make its work more legible
  • reduce trapped and hidden knowledge
  • connect decisions, workflows, systems, and context
  • create a structure AI can actually work with
  • improve contribution, reuse, and learning over time

This is not about buying a platform and hoping it solves the problem. It is about building the organisational layer that makes useful AI and better coordination possible.

Who should own it internally​

A KnowledgeFund should not belong to one isolated function alone.

The best internal starting group usually includes a mix of:

  • an executive sponsor
  • an operational lead
  • someone close to delivery or frontline work
  • someone with architecture or systems thinking ability
  • someone responsible for change, governance, or improvement

The goal is to combine authority, reality, and structural thinking.

A self-implementation sequence​

1. Choose one bounded starting area​

Do not start with the whole enterprise.

Start with:

  • one workflow
  • one business capability
  • one team
  • one recurring pain point
  • or one cross-functional handoff area

Choose somewhere that is already expensive in confusion, delay, duplication, or hidden dependency.

2. Map the real work​

Document the work as it actually happens.

Capture:

  • major steps
  • roles involved
  • systems touched
  • key artefacts
  • decision points
  • common delays
  • common rework
  • where people rely on memory, chat, or unwritten judgement

Do not model the idealised version. Model the real version.

3. Extract tacit knowledge​

Interview the people who actually make the workflow function.

Pull out:

  • rules of thumb
  • hidden exceptions
  • local workarounds
  • quality heuristics
  • escalation logic
  • recurring failure patterns
  • key dependency knowledge

This is where much of the real organisational capability is hiding.

4. Define the first ontology​

Create a structured way to describe the key things in the chosen area.

This usually includes:

  • goals
  • roles
  • workflows
  • tasks
  • systems
  • clients or stakeholders
  • artefacts
  • decisions
  • risks
  • dependencies
  • reusable assets
  • gaps

Without structure, knowledge will just accumulate into a larger mess.

5. Create the first connected knowledge layer​

Now start linking the pieces together.

For example:

  • workflows to roles
  • roles to systems
  • systems to artefacts
  • artefacts to decisions
  • decisions to goals
  • gaps to owners
  • reusable patterns to the conditions where they apply

This is what turns a repository into a living organisational model.

6. Embed it into live work​

Do not make the KnowledgeFund a side project or documentation tax.

It needs to be updated through real work.

That means:

  • decisions should leave traces in it
  • new lessons should be added during delivery
  • repeated work should pull from it
  • gaps should become visible while work is happening
  • contribution should be part of the workflow, not an afterthought

7. Use AI carefully as an interpreter and guide​

Once the structure is stable enough, use AI to help with:

  • classification
  • summarisation
  • suggested links
  • gap finding
  • duplication detection
  • surfacing relevant context
  • helping people contribute with less friction

AI should help the organisation use and maintain the KnowledgeFund. It should not be treated as the designer of the organisation.

8. Create contribution and reuse rules​

If the organisation wants this to live, it has to define what good contribution looks like.

For example:

  • what should be captured
  • what should be linked
  • what counts as a reusable asset
  • how gaps should be recorded
  • who reviews changes
  • how stale content is refreshed
  • how reused knowledge gets recognised

9. Review and expand​

Once the first slice starts working, review it honestly.

Ask:

  • what became clearer
  • what is still hidden
  • where the ontology is weak
  • where people still bypass the structure
  • what AI could now help with
  • what adjacent workflow should be next

Then expand into the next area using the same method.

What internal success should look like​

A first successful KnowledgeFund slice should leave the organisation with:

  • one domain that is more legible than before
  • less dependence on hidden individual memory
  • clearer connections between work, decisions, and systems
  • better visibility of gaps and dependencies
  • more reusable context for both people and AI
  • a repeatable method for the next domain

Common mistakes​

Avoid these traps:

  • trying to model the whole company too early
  • building a content dump without ontology
  • treating it as a documentation exercise only
  • making contribution feel like extra admin work
  • expecting AI to compensate for weak structure
  • leaving ownership vague
  • treating the first model as final instead of evolutionary

When outside help still makes sense​

Even if an organisation is doing this itself, outside support can still be useful for:

  • diagnostic challenge
  • ontology design help
  • workshop facilitation
  • implementation acceleration
  • governance design
  • capability coaching

But the long-term goal should still be internal capability, not permanent dependency.

The key mindset​

The most important shift is this:

Do not ask, "What AI platform should we buy to solve this?"

Ask:

  • where are we least legible to ourselves
  • where does work depend on trapped knowledge
  • what structure would make this part of the organisation easier to understand, reuse, and improve
  • how do we make that structure part of everyday work

That is how an organisation starts building its own KnowledgeFund.