AI needs a visible place in enterprise architecture
Introducing CEAF as a conceptual contribution from my graduate work, with a clear account of its limits.
CEAF reading series · Six short articles
CEAF: strategy, AI, and the enterprise
Six short articles connecting my Cognitive Enterprise Architecture Framework research with practical choices about AI strategy, delegated action, investment, and value.
Start with Part 1 ↗The series begins with my conceptual CEAF proposal, then develops its implications and related applications. The agentic AI articles extend the discussion using current primary guidance. CEAF has not been empirically validated through organisational implementation.
Read in order or choose the question most relevant to your work. Each article stands on its own.
Part 1 of 6
Part 2 of 6
AI strategy starts with the decisions that matter
Build an AI investment agenda around consequential decisions and the capabilities behind them.
Part 3 of 6
Decide when a workflow should become agentic
A practical way to decide whether a task needs an AI agent, a defined workflow, or simpler automation.
Part 4 of 6
Agentic AI needs an authority model
Treat delegated action as an organisational design decision, with explicit limits and accountable owners.
Part 5 of 6
Budget for the whole AI capability
Make the full cost and recurring ownership of an AI service visible before approving its expansion.
Part 6 of 6
Measure AI value at the level of work
Move AI evaluation from usage and generated output to the performance of the complete workflow.