Sign in Subscribe

Data Architecture

04
Aug
A white interlocking geometric pattern with sections missing from one corner, representing data infrastructure running on defaults where business decisions were never made.

Your data platform runs on decisions nobody made

The most consequential data infrastructure gaps are not the decisions that were made badly. They are the decisions nobody made — defaults the business inherited without signing off on.
3 min read
28
Jul
A white and blue glass skyscraper seen from below, floors stacked against a clear blue sky, representing the layered data infrastructure that AI workloads inherit from beneath.

The data infrastructure gaps that surface when AI moves in

When AI workloads arrive, they expose data infrastructure gaps that years of BI never surfaced. The problem is rarely the model — it is the data layer the model inherited.
4 min read
21
Jul
A close-up of a printed reference page showing a FORTRAN entry, representing the idea that precise definitions belong in code and configuration, not in the heads of individual data analysts.

The business metrics problem that data quality tools cannot fix

Business metric inconsistency is not a data quality problem. It is what happens when an organisation has never formally agreed on what its key metrics mean — and that is a governance gap, not a technical one.
4 min read
14
Jul
A close-up of black and grey metal pipes in an abstract industrial arrangement, representing the deliberate infrastructure choice that determines how data moves through the system.

Streaming is not the starting point. It is the upgrade.

Choosing streaming infrastructure before the business has asked for real-time data is not ambition. It is overhead.
4 min read
07
Jul
An architectural model alongside blueprint drawings pinned to a wall, representing the formal document that turns an implicit data agreement into an explicit one.

Data contracts are not the tests you run. They are the agreement you make.

A schema test catches a broken assumption after the fact. A data contract prevents the assumption from being broken in the first place.
3 min read
30
Jun
A grayscale close-up of interlocking steel building frames casting angular shadows, representing the structural sequencing that a data platform migration depends on.

Data platform migrations rarely fail on tooling. They fail on sequencing.

A data platform migration is a sequencing problem. Most teams start at step two.
4 min read
23
Jun
A grayscale close-up of an exposed metal structural frame with geometric grid of beams in stark contrast, representing the modular architecture that keeps each platform decision independent.

The data platform decision that comes before build or buy

The build vs buy decision is not the first question. The first question is whether any third-party data platform component is needed at all, and most teams skip it.
4 min read
09
Jun
A black and white close-up of a cracked concrete surface with fracture lines across the frame, representing the silent structural failure that schema drift introduces at the data layer.

Schema drift is a business risk. Most teams treat it as a technical detail.

A schema change is not just a technical event. When it goes unmanaged, it silently degrades data quality, breaks downstream systems, and makes AI features unpredictable in ways that look like model problems — not data problems.
5 min read
02
Jun
A close-up of a fractured concrete wall with a prominent crack running through weathered grey stone, representing the structural failure that occurs when data pipelines lack clear ownership.

Your data pipelines aren't failing because of bad code

Most data pipeline failures at scale are not bad code or data infrastructure problems. They are ownership problems — and until the contract between data producers and data consumers is made explicit, the same failures will keep repeating.
4 min read
26
May
Angular concrete forms and dramatic lighting inside a brutalist structure, representing the deliberate design of spaces where different roles can collaborate and build toward a common goal.

You hired the right people. The right structure is next.

Getting the right people on your data team is only half the work. The structure around them — how they communicate, who leads them, how they build shared understanding — determines whether that investment compounds or quietly erodes.
4 min read