How an organisation would build its own KnowledgeFund
Most organisations do not have a technology problem when it comes to AI.
They have a context problem.
The knowledge of how the organisation really works is scattered across people, systems, habits, exceptions, workarounds, and fragments of documentation. Important decisions are often poorly connected to the workflows they affect. Teams compensate socially. They ask around. They rely on memory. They improvise.
That is survivable in a human-only organisation. It becomes much more visible when AI enters the picture.
AI quickly exposes whether the organisation has a coherent enough internal structure to support shared understanding, guided action, and repeatable improvement.
That is why more organisations will need to build something like a KnowledgeFund for themselves.
Not as a document repository. Not as a wiki with a new name. Not as a one-off architecture exercise.
But as a structured, living organisational knowledge system that helps the company learn, guide work, reuse what it knows, and become more capable over time.
So how would an organisation actually do it?
Start smaller than your ambition
The first mistake would be trying to model the whole company at once. That sounds strategic, but it usually produces abstraction before usefulness.
A KnowledgeFund should begin in a bounded domain where real work is already happening and where the pain of fragmentation is visible.
That might be:
- a delivery workflow
- a client onboarding process
- a sales-to-delivery handoff
- a tender response process
- a recurring operational area where too much knowledge currently lives in people's heads
The point is not to create a perfect enterprise model on day one. The point is to start where real work, real dependency, and real ambiguity already exist.
Map the work as it actually happens
Once a starting territory is chosen, the organisation has to map the work as it really happens.
Not the cleaned-up version. Not the process slide. Not the compliance version.
The real version.
What steps are actually taken? Who is involved? Which systems are touched? Where are decisions made? Where are delays introduced? Where do people rely on hidden knowledge, informal chat, memory, or "ask the person who knows"?
This matters because the raw material of a KnowledgeFund is not abstract theory. It is the lived operational reality of the organisation.
Extract tacit knowledge
This is one of the most important parts.
Most organisations run on a huge amount of unwritten judgement:
- rules of thumb
- exceptions
- escalation instincts
- quality heuristics
- client-specific knowledge
- system quirks
- local workarounds
Often the most important knowledge is not absent, it is trapped. It exists, but only in forms that do not travel well.
Building a KnowledgeFund means pulling that knowledge into a more shareable and structured form.
Not flattening it. Not pretending everything is simple. But making it more visible, more connected, and less dependent on individual memory.
Create an ontology
That word can sound heavier than it needs to be.
In practice it means creating a structured way to describe what exists in the business and how those things relate.
For example:
- goals
- roles
- workflows
- clients
- systems
- artefacts
- decisions
- risks
- dependencies
- reusable assets
- gaps
A KnowledgeFund needs this kind of backbone so that knowledge does not just accumulate as another pile of disconnected content.
This is one of the key distinctions.
A normal knowledge repository stores information. A KnowledgeFund gives the organisation a way to place knowledge into a meaningful structure.
That structure is what makes the company more legible to itself.
Build a shared knowledge layer
Once the ontology exists, even in early form, the organisation can start building a shared knowledge layer around it.
This means connecting:
- workflows to roles
- roles to systems
- systems to artefacts
- artefacts to decisions
- decisions to goals
- gaps to owners and actions
- reusable patterns to where they should be applied
Now the KnowledgeFund starts becoming more than documentation. It becomes an organisational model.
Embed it into live work
This is where many attempts fail.
If contributing to the system feels like extra admin work after the "real work" is done, it will slowly die.
A KnowledgeFund has to be embedded into live work.
People should update it through delivery, not through a separate clerical ritual. Decisions should leave traces in it. New patterns should be added through use. Reuse should pull from it naturally. Gaps should become visible at the moment they matter.
The system has to participate in the work of the organisation, not sit outside it.
Use AI to help interpret and connect
This is also where AI starts becoming genuinely useful.
Not because AI magically solves organisational design, but because it can help interpret, classify, connect, summarise, and surface knowledge at much greater scale than traditional manual systems.
Used well, AI can:
- help staff contribute with less friction
- suggest where new knowledge belongs
- point to missing context
- detect duplicated patterns
- highlight stale logic
- surface likely gaps
- bring relevant context forward when work is happening
Used badly, AI just accelerates confusion.
The difference is whether the organisation has built enough structure for AI to work with.
Reward contribution and reuse
If the organisation wants the KnowledgeFund to grow into something real, it has to value more than visible frontline performance.
It has to reward people who:
- make the business more legible
- identify gaps
- improve shared clarity
- create reusable assets
- strengthen handoffs
- reduce dependence on hidden expertise
Many organisations accidentally reward local heroics while under-rewarding the work that makes the whole system better.
A KnowledgeFund only works if the organisation starts treating shared intelligence as part of the job.
Expand gradually
Once the first slice is working, the organisation can extend the model into adjacent workflows, teams, and knowledge domains.
The ontology gets richer. The reuse patterns get clearer. The contribution model matures. AI assistance improves because the surrounding context improves.
Over time, the company builds not just a better knowledge base, but a stronger ability to learn, adapt, and act with shared context.
That is the deeper point.
A KnowledgeFund is not just a repository of what the company knows. It is a way for the company to become more coherent, more teachable, more reusable, and more capable of working with AI without becoming dependent on chaos, tribal memory, or vendor narratives alone.
The right first question
So if an organisation wants to create a KnowledgeFund for itself, it should not ask, "What platform should we buy?"
It should ask:
- where does our real work currently depend on trapped knowledge
- where are we least legible to ourselves
- what would it take to make that part of the organisation more structured, shared, and reusable
- how do we turn that into a living system, not just a documentation exercise
That is where the work begins. And that is probably the right place to start.