The cross-functional team determines the type of AI solution by asking how end-users will use or interact with it, whether the same outcome is needed every run, how accurate it must be, how far the model must be explained to end-users, how long it may take to produce results, whether the training data reflects the real-world situations it will address, and which EU AI Act risk category the use falls under.
The graph holds this control, the 0 it maps to, and the evidence behind each claim, over MCP and REST.