What is agent orchestration?
LLM & AILevel: IntermediateAgent Architectures
What is agent orchestration?
Hints
- Identify the stage: pretraining, adaptation, generation or evaluation.
- Name the inputs, constraints and a case where the approach fails.
Answer
Orchestration coordinates the flow of tasks across agents, tools, and human interactions:
- Deciding which agent to call based on the task type
- Passing outputs from one agent as inputs to the next
- Managing state shared across agents
- Handling errors — if one agent fails, decide whether to retry, fallback, or escalate
- Budget tracking — enforce token/cost limits across the pipeline
Implementation options:
- LangGraph — graph-based stateful workflows; nodes = agents/tools, edges = transitions
- Prefect / Airflow — DAG-based orchestration for less dynamic workflows
- Custom orchestrator — LLM-based supervisor that reads task state and decides next action
Source and licence: JustInCache & contributors — Awesome AI Interviews · Apache-2.0