ENTERPRISE ARCHITECTURE · DATA ARCHITECTURE · DESIGN DECISIONS
Architecture is the discipline of making important choices visible. It connects business intent to the capabilities, processes, information and technology needed to make that intent real.
Too often, architecture is reduced to diagrams, standards or a description of the target state. Those things can be useful, but they are not the work itself. The real work is understanding the problem, exposing the dependencies, making trade-offs explicit and helping people take decisions that the organisation can live with.
I approach architecture as a practical discipline. It should help an organisation move forward without losing coherence. It should create enough direction for teams to act independently, while making sure that local decisions do not quietly undermine the bigger picture.
Architecture starts with the decision
A useful architecture conversation begins with a question: what are we trying to change, enable or protect? The answer might concern a business capability, a customer process, a regulatory obligation, a product lifecycle or the information needed to make a decision.
Only then can we decide which architectural concerns matter. Sometimes the issue is an application boundary. Sometimes it is a data ownership problem, a missing business capability, an integration dependency or an operating model that no longer fits the way the organisation works. Starting with a preferred technology or framework too early risks solving the wrong problem very elegantly.
The architecture perspectives
Enterprise architecture
Connecting strategy, capabilities, value streams, processes, organisation and technology. The purpose is not to model everything. It is to understand the relationships and dependencies that matter when the organisation changes.
Data architecture
Designing how information is created, owned, integrated, governed and used. This includes data platforms and flows, but also the responsibilities and business processes that determine whether information is reliable and fit for purpose.
Semantic architecture
Making business meaning explicit and reusable. Canonical models, semantic layers, ontologies and knowledge graphs help establish what concepts mean, how they relate and where different systems need to share an understanding.
Solution and technology architecture
Turning the intent into implementable choices: application responsibilities, integration patterns, platforms, security, deployment and operational constraints. A sound design has to work in the organisation that will build, run and evolve it.
A few principles I keep coming back to
- Start with the business consequence. Architecture should be anchored in a real outcome, risk or constraint.
- Make decisions explicit. Record the reasoning and trade-offs, not just the selected technology.
- Respect the existing landscape. A target state is only credible when there is a practical path from today’s reality.
- Design for change. Boundaries, ownership and contracts should make evolution possible without requiring everything to change at once.
- Keep governance useful. Standards and review mechanisms should improve decisions, not become work products in their own right.
Selected perspectives
The Structure Behind Data-Driven Decisions
Why becoming data-driven depends on more than dashboards and platforms.
Data Architecture Patterns: Decisions for the AI Era
Looking beyond pattern names to the decisions and trade-offs behind them.
The Semantic Layer: Let’s Dive In
Making business meaning reusable across data and analytical products.
Your Silver Layer Is Not Missing Data. It Is Missing Decisions.
Why architecture needs to encode decisions, not simply transform data.
Concept of Canonical Models
Shared structures as a way to reduce translation and ambiguity.
Architecture in practice
Good architecture is not the pursuit of theoretical perfection. It is the deliberate management of complexity. It gives people a clearer view of the consequences of their choices and a shared basis for moving forward.
That is the standard I find useful: not whether an architecture looks complete, but whether it helps the organisation make better decisions and deliver change with confidence.
For the information and ownership side of this work, explore Master Data. For the implications of AI on enterprise meaning and decision-making, visit AI & Intelligence. You can also browse the full article archive.