Defines SLIs, SLOs and an error-budget policy from a service's user journeys, with multi-window burn-rate alert rules. Use when alerting is noisy or reliability targets are vague.
