Canonical Scenarios
Canonical scenarios represent common, repeatable production patterns observed across real teams — not demos or idealized workloads.
Usage pattern stability indicates how consistent real-world implementations of a scenario tend to be across teams.
AI Agent
Variable usage patternMulti-step execution with tool calls and retries. Cost scales with execution depth.
Dominant driver: Execution depth
Customer Support
Stable usage patternHigh-volume conversational workload with predictable traffic.
Dominant driver: Request volume
Internal Copilot
Variable usage patternKnowledge work assistant with longer prompts and habit-driven usage.
Dominant driver: Habit intensity
AI Search (RAG)
Variable usage patternRetrieval-heavy workloads dominated by injected context tokens.
Dominant driver: Context size