Agentic Architecture & Orchestration
Turn “have the AI do a big job” into a system that survives production.
The largest domain, because it is where things break in production. A weak prompt gives a weak answer; a weak architecture double-charges a customer, loops forever, or burns a budget overnight. It covers how to turn a goal into a reliable system: decompose the work into named steps, checkpoint state so a run can resume, make every step idempotent so retries are harmless, compose many agents into an orchestrator → doer → validator org chart, and bound the risky edges with guardrails that fail closed and autonomy granted in proportion to risk.
7 task statements on the blueprint
- 1.1Design and implement agentic loops for autonomous task execution4 objectives · 5 exam traps · 3 primary sources · 14 practice questions
- 1.2Orchestrate multi-agent systems with coordinator-subagent patterns4 objectives · 5 exam traps · 2 primary sources · 16 practice questions
- 1.3Configure subagent invocation, context passing, and spawning4 objectives · 6 exam traps · 2 primary sources · 12 practice questions
- 1.4Implement multi-step workflows with enforcement and handoff patterns5 objectives · 5 exam traps · 2 primary sources · 17 practice questions
- 1.5Apply Agent SDK hooks for tool call interception and data normalization4 objectives · 6 exam traps · 2 primary sources · 6 practice questions
- 1.6Design task decomposition strategies for complex workflows4 objectives · 5 exam traps · 2 primary sources · 16 practice questions
- 1.7Manage session state, resumption, and forking4 objectives · 5 exam traps · 2 primary sources · 10 practice questions
The other domains
Working the whole blueprint? The interactive study guide carries the concept cards, the mock quiz, the audio lessons and the practice simulation.