The publication
The Applied Layer research programme is organised around the five disciplines that decide whether enterprise AI reaches production value.
Each pillar keeps the same Reader, Member, and Patron access pattern: public orientation, member briefings and excerpts, and patron access to the underlying research reports with citations preserved.
Pillar 1 of 5 · 2 pieces
The foundation study for the five-pillar programme — why the applied layer shapes enterprise AI outcomes, what the evidence establishes, what it recalibrates, and what it leaves open for the field interviews to test.
The most consequential layer of the AI buildout is not the foundation models themselves, but what sits between them and the organisations that deploy them: architecture, integration, evaluation, and governance.
Pillar 2 of 5 · 2 pieces
Patterns that distinguish production from demo — why the architecture wrapped around the model, not the model itself, decides enterprise AI outcomes.
Central proposition.
Pillar 3 of 5 · 2 pieces
Why the operating model — not the technology choice — decides enterprise AI outcomes, and how to choose the structure that fits.
Central proposition.
Pillar 4 of 5 · 2 pieces
What production actually costs in 2026 — and which platform fits your workload
The headline cost of model inference is a small and shrinking fraction of what enterprises actually spend to run generative AI in production.
Pillar 5 of 5 · 2 pieces
Evaluation as practice, governance as delivery — the two disciplines that decide whether enterprise AI earns operational trust
Enterprise AI systems fail in production in characteristic ways.
