You are releasing a new orchestrator prompt and tool schema for a production multi-agent concierge. Risk of quality regression is moderate; you need limited real traffic exposure with fast rollback. Which release methodology fits best?
Select an answer to reveal the explanation.
Short Explanation
Answer B. Deploying multi-agent solutions includes release methodologies such as DTAP, blue/green, and canary. Canary limits blast radius while gathering real signals, with rollback if metrics tank. 100% big-bang without rollback is risky. Laptop-as-prod and unversioned live edits fail operational standards.
Full Explanation
Correct answer: B. Choose canary/progressive delivery for moderate-risk agent changes: route a small cohort, compare evaluation and operational KPIs against the baseline version, then promote or roll back. Keep artifacts versioned (prompts, graphs, tool schemas).
A is incorrect: all-at-once cutover maximizes user impact on regression.
C is incorrect: developer machines are not resilient multi-tenant production platforms.
D is incorrect: unversioned manual edits break auditability and safe rollback.
Blue/green is also valid when you need instant environment swap; canary emphasizes percentage-based exposure.