This course contains the use of artificial intelligence. This is Part 1 of a three-part programme on AI-native platform engineering. It covers the decision and operating layer and the execution substrate: modules 1 to 5 and module 13 of the full curriculum. Part 2 covers identity, security, supply chain, evaluation and observability. Part 3 covers cost, governance, human oversight and the capstone. Each part stands on its own and produces its own deliverables, and only the full programme answers all seven control planes.
The problem this part solves
Your organisation has thirty or forty agent prototypes, no way to say which of them should exist, no named owner for most of them, and a board asking what the return is. The prototypes that do work run on one laptop, in one trust context, for one user, with unlimited patience and no cost ceiling. Production has none of those properties, and most teams discover which parts of the demo were load-bearing only after the first incident.
What you will learn
The first half is the operating model. You will run a suitability assessment that rejects agentic solutions to deterministic problems, derive autonomy from reversibility and blast radius rather than from enthusiasm, cost the failure before you cost the build, evaluate managed platforms against seven control planes instead of feature lists, and stand up an intake process where every agent has a charter, a named owner and kill criteria agreed before launch. You will design the platform team, make the funding case, find your shadow AI estate and migrate it without stopping delivery.
