Definition
Orchestration determines what happens before and after a model call: which context is retrieved, which tools can run, what must be checked, and how failure is handled.
Why it matters
A capable model without good orchestration produces inconsistent work; orchestration turns it into a system.
Business example
A request is classified, enriched from the CRM, drafted by AI, checked against policy, then routed to the right queue.
When to use it
Use it when an AI experience has more than one source, tool, or decision point.
When not to use it
Do not overbuild it for a narrow, single-prompt assistive task.
How Automathing approaches it
We make the flow explicit, observable, and recoverable before expanding capabilities.