01
What is it?
The entity: what exists in the business, what distinguishes it from everything else, and how it is identified.
Ontology Engineering
Ontology engineering: making the business machine-readable.
An enterprise ontology is the explicit, versioned, owned statement of what the business knows and how it relates. It is the contract between human intent and machine reasoning.
What is ontology engineering at Cyberwisdom?
Ontology engineering at Cyberwisdom Group is the discipline of making an organisation's business meaning explicit, structured and machine-usable, and then maintaining it as a governed asset. The group applies its <b>HAKDB</b> model — Human, AI, Knowledge, Data and Business Oriented Ontology — using the <b>R-LED-R</b> framework (Rules and Logical based Entity, Description and Relationships) as the business and technical baseline for scaling AI application accuracy. The capability is led by Origin Big Data, a group company, and connects directly to the group's knowledge management and competence systems. In practice it covers entity and relationship modelling, business rule extraction, capability graph construction, change control with named ownership, and the evaluation datasets that make AI reliability measurable against the organisation's own truth.
What an ontology is
Five questions an enterprise ontology must answer.
If a model cannot answer these five questions about a concept, it is guessing — and a model that guesses is not an enterprise system.
02
What is true of it?
The description: its attributes, their types, their validity, and the conditions under which each attribute holds.
03
How does it relate?
The relationships: what connects to what, with what cardinality, and under what conditions.
04
What rules govern it?
The logic: the conditions that determine validity, eligibility, authority and required action — the part most often left implicit.
05
Who owns it?
The accountability: which team owns the definition, approves changes, and is answerable when it is wrong.
The models
HAKDB and R-LED-R.
Cyberwisdom's own approach uses two constructs — one to frame the ontology, one to engineer it.
HAKDB — the framing model
Human, AI, Knowledge, Data and Business Oriented Ontology. The point of HAKDB is that AI is not added as a separate layer on top of the enterprise — it is modelled within the same frame as the humans, the knowledge, the data and the business rules that already constitute the organisation.
- Human — roles, judgement, accountability
- AI — agents, models, authority limits
- Knowledge — institutional memory made explicit
- Data — systems of record
- Business — the objectives and rules that govern it
R-LED-R — the engineering framework
Rules and Logical based Entity, Description and Relationships Framework. The baseline used to define capabilities and scale AI application with accuracy — applied before and during implementation rather than retrofitted after an error.
- Rules — explicit conditions and validity
- Logical — formal and testable structure
- Entity — the things the business knows
- Description — what is true of them
- Relationships — how they connect and constrain
Method
How an ontology is actually built.
This is a social process as much as a technical one. The disagreements you find are the value.
STEP 01
Scope
Choose a business domain that matters and is bounded enough to finish. A complete ontology of everything is a project that never completes.
STEP 02
Elicit
Extract meaning from the people who hold it: business owners, operational experts, and the artefacts they leave behind in policies, systems and decisions.
STEP 03
Reconcile
Surface and resolve the disagreements. Where two teams define the same concept differently, that conflict is a live business risk — the ontology makes it visible.
STEP 04
Model
Encode entities, relationships, attributes and rules with named ownership and version control.
STEP 05
Ground
Connect the ontology to source systems and retrieval, so grounding becomes structural rather than a prompt instruction.
STEP 06
Evaluate
Encode real, permissioned cases against the model, so AI accuracy can be measured against business truth rather than impression.
Capability graph
Ontology expressed as what the business can do.
From a data model to a capability graph
Cyberwisdom's capability architecture links what the organisation knows, what it is permitted to do, and what it is actually able to do. That is a different and more useful question than what data it holds.
- <b>Knowledge nodes.</b> The concepts, policies, procedures and decision rules the organisation has formalised.
- <b>Capability nodes.</b> The business capabilities those knowledge elements enable — a process the organisation can execute, with the conditions attached.
- <b>Authority edges.</b> Which roles and systems may execute which capabilities, and under which approvals.
- <b>Gap visibility.</b> Where a capability is defined but not evidenced, or evidenced but not documented — the two most common sources of operational risk.
- <b>Agent-ready structure.</b> A graph that an agent can resolve against is what makes grounded, explainable enterprise AI possible at scale.