EDITORIAL REF.
IInsights
Cabinet insights - on agentic architecture and AI governance.
An editorial production built over time. The tone: practitioner editorial, no overpromise, no unnecessary jargon. The frequency: what deserves to be written, not what cadence demands.
01Featured cabinet publication
Pillar article - recommended entry point.
An original publication, written through the cabinet lens, to read first to understand the practice's perspective. Other formats will arrive at the pace of the editorial program - no volume race.
Other publications
Notes and support articles published as they come.
- REF.01EU AI Act & governanceMay 27, 2026
The EU AI Act for non-regulated companies: what really changes
A cabinet reading of Regulation (EU) 2024/1689 for companies that think they are not concerned: application timeline, risk levels, transparency, GDPR overlap and five architecture reflexes to prepare.
- REF.02Agentic architectureMay 27, 2026
Agentic architecture in enterprise: what really changes
Why enterprise AI should be treated as an operating layer, not as a set of isolated assistants. Five structural differences between an AI agent and a governed agentic system.
02Editorial status
A short library, but a governed one.
This page does not pretend the English corpus is already mature. It separates cabinet publications, historical archives and upcoming topics so the signal remains clear for leaders, search engines and LLMs.
01
Published
Cabinet articles are published when they are sourced, dated and connected to the method, architecture, sovereignty or service pages.
02
In preparation
The English layer follows the priority clusters: agentic architecture, EU AI Act, sovereignty, ROI and adoption.
03
Archives
Historical French-only content remains accessible from the archive, but is kept separate from the current cabinet corpus.
03Editorial program
Five thematic clusters.
The cabinet editorial revolves around five angles, chosen for their enterprise relevance and stability over time. The program is deliberately tight - each cluster grows when a topic deserves to be addressed.
- REF.01
Agentic architecture
Patterns, anti-patterns, stack choices, field returns on enterprise agent deployments.
- REF.02
EU AI Act & governance
Operational readings of the European regulation, concrete translation into architecture, anticipation of evolutions.
- REF.03
ROI & prioritization
Evaluation methods, prioritization matrices, business value measurement of AI systems.
- REF.04
Sovereignty & European stack
Sovereign models, European hosting, stack choices, Mistral and the French ecosystem.
- REF.05
Transformation & adoption
Change management, team training, human-in-the-loop governance, agentic project leadership.
Stay informed
Receive upcoming publications.
No newsletter at this stage - publications will appear on the site and usual professional channels (LinkedIn notably, without overload). For a direct discussion on any of the themes above, an exchange is always possible.