You need a multi-agent claims system where a coordinator assigns work to fraud, coverage, and settlement specialists, aggregates results, and returns one decision package. Specialists should not freely call each other. Which orchestration pattern should you implement?
Select an answer to reveal the explanation.
Short Explanation
You want hub-and-spoke/orchestrator-subagent. A coordinator routes, scopes, and aggregates while specialists stay in their lanes, giving you one decision package without uncontrolled chatter.
Full Explanation
Correct pattern — hub-and-spoke (orchestrator-subagent): Multi-agent orchestration patterns include hub-and-spoke, sequential, parallel, peer-to-peer, and orchestrator-subagent. Claims-style coordination with scoped fraud, coverage, and settlement specialists and a single aggregated package maps cleanly to hub-and-spoke / orchestrator-subagent: a coordinator routes tasks, specialists execute within scope, and the hub aggregates outputs into one decision package.
Why the fully peer-to-peer mesh with no coordinator is wrong: Unconstrained peer mesh complicates control, audit, and scope, and lets specialists message each other freely.
Why the strict sequential pipeline only is wrong: Forced sequencing underuses parallelism when fraud and coverage checks are independent.
Why unbounded agent spawning with no aggregation step is wrong: Spawning without aggregation fails the one decision package requirement.
Exam tip: Match pattern to control needs: hub when aggregation and routing dominate.