0 alternative definitions
An engineering practice for designing the execution system surrounding AI models, including context and state management, tool-use orchestration, guardrails, and feedback loops, to enable an AI agent to perform tasks reliably.
