A support solution needs a coordinator agent that classifies intent, then routes work to billing, technical, or shipping specialist agents. Specialists should not call each other directly; only the coordinator consolidates results for the user. Which multi-agent orchestration pattern should you implement?
Select an answer to reveal the explanation.
Short Explanation
Answer D. When a single coordinator owns routing and synthesis and specialists stay isolated from each other, you want hub-and-spoke / orchestrator-subagent orchestration. That matches intent classification plus specialist dispatch. Full mesh peer-to-peer creates uncontrolled lateral chatter. Parallel fan-out without aggregation leaves the user with fragmented outputs. A fixed sequential pipeline wastes work when only one specialty is needed.
Full Explanation
Correct answer: D. Multi-agent orchestration patterns on AI-500 include hub-and-spoke, sequential, parallel, peer-to-peer, and orchestrator-subagent. The described control flow—central classification, exclusive coordinator mediation, specialist isolation, and consolidated user response—is classic hub-and-spoke / orchestrator-subagent.
A is wrong because peer-to-peer allows direct specialist-to-specialist communication, which the requirement forbids and which complicates governance and tracing.
B is wrong because unordered parallel work without a merge step does not produce a single coordinated answer.
C is wrong because always running every specialty is inefficient and ignores intent-based routing.
Implementation notes: use Agent Framework or LangGraph to encode the hub graph, preserve correlation IDs across spoke calls, and keep tool scopes per specialist.