Data Engineering and Business Ontology
The semantic layer: definitions, entities, identifiers, rules and permissions that agents actually need.
What this collection argues
Explain the engineering under the agent — shared business context, ontology as an enterprise language, the difference between data access and usable context, and a readiness assessment that can be run honestly.
In this collection
5 articles. Each one opens with the business problem, names the executive who owns the outcome, and ends with the metrics that would show whether the work paid.
Why AI Cannot Fix Data Problems That the Business Has Never Defined
An AI system applied on top of undefined terms, unowned records and duplicated entities will reproduce the ambiguity at scale; only the business can supply the definitions the model then depends on.
From Fragmented Enterprise Data to a Shared Business Context
Connecting systems, records, identifiers, relationships and definitions into one shared business context is the prerequisite for any agent that acts, and it is built entity by entity.
Business Ontology Explained: Connecting Data to How a Company Actually Works
A business ontology is the explicit set of entities, relationships, events, rules and permissions a company uses to describe itself; without it an agent can retrieve data but never understand the business.
Why an AI Agent Needs More Than Database Access
A schema tells an agent where data lives; what it needs is context, authorisation, business rules, tools, validation and bounded execution — the difference between a query interface and a governed actor.
The Enterprise Data Readiness Checklist for AI Projects
A readiness assessment across availability, quality, ownership, integration, access and auditability that produces a defensible go or no-go, rather than a score assembled from optimism.