Category: Data Management
-
The layer has a name: Ontology.
Traditional BI answers “๐ช๐ต๐ฎ๐ ๐ต๐ฎ๐ฝ๐ฝ๐ฒ๐ป๐ฒ๐ฑ?” Semantic models answer “๐๐ผ๐ ๐๐ต๐ผ๐๐น๐ฑ ๐๐ต๐ฒ ๐ฏ๐๐๐ถ๐ป๐ฒ๐๐ ๐บ๐ฒ๐ฎ๐๐๐ฟ๐ฒ ๐ถ๐?” Taxonomies answer “๐ช๐ต๐ฎ๐ ๐ธ๐ถ๐ป๐ฑ ๐ผ๐ณ ๐๐ต๐ถ๐ป๐ด ๐ถ๐ ๐๐ต๐ถ๐?” Knowledge graphs answer “๐ช๐ต๐ฎ๐ ๐ถ๐ ๐ฎ๐ฐ๐๐๐ฎ๐น๐น๐ ๐ฐ๐ผ๐ป๐ป๐ฒ๐ฐ๐๐ฒ๐ฑ, ๐ฟ๐ถ๐ด๐ต๐ ๐ป๐ผ๐?” ๐ข๐ป๐๐ผ๐น๐ผ๐ด๐ถ๐ฒ๐ ๐ฎ๐ป๐๐๐ฒ๐ฟ “๐๐ผ๐ ๐ฎ๐ฟ๐ฒ ๐ฎ๐น๐น ๐ฏ๐๐๐ถ๐ป๐ฒ๐๐ ๐ฐ๐ผ๐ป๐ฐ๐ฒ๐ฝ๐๐ ๐ฐ๐ผ๐ป๐ป๐ฒ๐ฐ๐๐ฒ๐ฑ?” Your AI agent doesn’t know what a customer is. It knows you have a table called…
-
Data Models: Conceptual vs Logical vs Physical
Every database starts as an idea. By the time it’s running in production, that idea has passed through three distinct stages, each one adding detail the last one didn’t have. Skipping straight to the last stage is how projects end up with schemas that technically work but don’t actually reflect how the business thinks. Understanding…
-
Your Silver Layer Is Not Missing Data. It Is Missing Decisions.
Most Silver layers are full of cleaned tables, standardised formats and automated transformations. But that is not where the real value sits. The real value of Silver is that it turns business decisions into reusable enterprise meaning. That is why a canonical model is not just a shared schema between applications. It is the organisation’s…
-
AI Governance: A Structural Perspective for Practitioners
As an architect, I’m watching a still-immature space, one where organisations are throwing themselves into AI faster than the structure around it can keep up. I keep seeing the same moment play out, usually right after something goes slightly wrong: who was supposed to be watching this? A chatbot gives a customer bad advice. A…
-
The Structure Behind Data-Driven Decisions
Every organisation that claims to be “data-driven” is really making a bet on a structure it rarely talks about: the data architecture. It’s the reason a dashboard shows the right number, the reason a report doesn’t break when a new product line launches, and the reason two teams looking at “revenue” actually mean the same…
-
Clear roles. Stronger governance. Better data.
Data governance often fails not because organizations lack policies or tools, but because nobody is quite sure who is supposed to do what. A data quality issue surfaces, and three teams point at each other. A new data product needs an owner, and the request stalls in email threads. A security exception is requested, and…
-
Where Medallion Architecture and Canonical Models Actually Meet
Medallion architecture has become the default mental model for organizing a data platform: land raw data in Bronze, clean and conform it in Silver, serve trusted, business-ready data from Gold. It’s simple, it’s intuitive, and it works, right up until three teams build three different “customer” tables and nobody can agree on the number in…
-
Distributed Attribute Ownership Specification – Template
What this describes:ย how ownership/responsibility for a single data attribute is distributed across multiple parties, and how that responsibilityย hands offย as the attribute (or its governing lifecycle) moves through states. Fill in every โน โบ placeholder. Delete the guidance notes when done. 1. Identity Field Value Attribute โนname of the attribute being governedโบ Type / shape โนe.g.…
-
Data Contract – Template
How to use this.ย A data contract is an agreement between the people who produce a dataset and the people who consume it. It exists to replace assumption with accountability, not to generate paperwork. Fill the Minimum Viable Contract (sections 1โ5) for every governed dataset. Add sections 6โ9 only when the data’s scale, risk, or number of consumers…
-
From Data Responsibility to Trust
Let us get real. Most companies talk a big game about data platforms, self-service analytics, and unlocking the potential of data. Then complicated topics like PII enter the scene, creating chaos or prompting many to turn a blind eye to the problem. This is why the role of a data platform owner is critical. Without…