When designing intelligent agents, understanding the distinction between long-running orchestrations and short-lived tasks becomes a critical architectural decision that shapes how your agent will manage complex workflows versus quick interactions. As an Agent Designer, you must carefully balance the need for persistent state management in extended processes against the efficiency of stateless, rapid executions—a choice that ultimately defines how your agent will perceive time and maintain context throughout its operational lifecycle.

Short-Term Tasks

A short-lived task in the context of orchestration refers to a process or operation designed to execute quickly and typically synchronously, completing within a relatively short time frame.

Key Characteristics:

Examples:

Why Use Short-Lived Tasks?