How do you design a customer deployment so their team can run it without you, and you make yourself obsolete?

This is a values-and-judgment question disguised as a systems question. The signal is whether the candidate treats observability and runbooks as deliverables on the same footing as the feature, and whether they can describe a concrete handoff bar (the customer's on-call resolves a defined incident class without paging the FDE). Weak answers describe great monitoring for themselves. Probe with 'how do you know the handoff actually worked' to force a measurable definition, and 'the customer wants you to stay and own ops' to test whether they push back on becoming a permanent dependency.