Insights
Field notes on agentic enterprise architecture: where AI should do architecture work, how to check its answers, and how to govern AI agents across platforms.
Latest · Estate hygiene and regulation
Application rationalisation with AI: what to automate and what stays a judgement
Application portfolio rationalisation with AI: which steps to automate, which stay human calls, and a six-step TIME method a steering committee can check.
Agentic enterprise architecture
Where AI should do architecture work, and what changes for EA teams.
Agentic transformation: the EA and planning tooling landscape as of 2026
How AI agents are changing enterprise architecture, transformation planning and organisational change in 2026, and how to choose where the work runs.
What changes in the EA tool when agents do the work
EA tools are turning into the governed context that agents read. What moves out of the tool, where that work lands, and what EA teams should do now.
A general AI assistant with your EA tool's MCP server isn't an EA agent
A connected assistant handles one record well. Estate questions need scale, identity across systems and checked figures. How to tell which you need.
The problem with AI inside EA tools
AI built into an EA tool sees one repository, is priced for seats rather than agent work, and can’t show its sources. Five limits, and eight questions to ask.
Trust and proof
How to check an AI answer about your estate before it reaches a forum.
AI-agent governance
Governing agents across platforms the way EA governs applications.
Estate hygiene and regulation
Reconciling systems of record, and drafting regulatory registers from the estate.
CMDB vs EA repository: why they disagree and how to reconcile them
The CMDB records what runs; the EA repository records what the business has and plans. Why they disagree, and a method to reconcile them so both can be trusted.
How to build an EU AI Act AI inventory from your enterprise architecture
Build the EU AI Act AI inventory from your EA repository, CMDB, procurement and agent platforms: the fields each entry needs, risk classification and upkeep.