What is Multi-agent orchestration?
Multi-agent orchestration is the coordination of several specialised AI agents on one task, each handling part of it under a supervisor that can halt the whole run.
Also called: agent swarms, multi-agent systems
In more detail
Work is divided between agents with narrower responsibilities — one retrieves, one drafts, one checks — either in sequence or in parallel. A supervising process tracks progress, resolves conflicts and stops execution when something goes wrong.
Why it matters
Narrow agents are easier to evaluate, cheaper to run and simpler to constrain than one agent asked to do everything. The supervisor is also the natural place to enforce limits and capture the audit trail.
How iLeaf approaches it
iLeaf builds both sequential and swarm topologies, chosen by how independent the sub-tasks genuinely are, with the supervisor wired into the Governor Engine so a run can always be stopped.