Glossary
Key concepts in AI agent operations — clear definitions, practical context, and links to deep-dive guides.
AgentOps
AgentOps (Agent Operations) is the discipline of building, deploying, observing, and managing AI agents in production. Learn the core principles, key practices, and how it differs from MLOps and DevOps.
AI Agent Deployment
AI agent deployment is the process of moving an AI agent from development to production where it serves real users. Learn the strategies, risks, and operational requirements for safe agent deployment.
AI Agent Evaluation
AI agent evaluation is the systematic process of measuring whether AI agents work correctly before and after deployment. Learn the methods, metrics, and frameworks that matter.
AI Agent Guardrails
AI agent guardrails are safety constraints that prevent agents from taking harmful, unauthorized, or out-of-scope actions in production. Learn the types, how they work, and why every production agent needs them.
AI Agent Lifecycle
AI agent lifecycle management is the end-to-end practice of managing an AI agent from initial development through production operation, iteration, and retirement. Learn the stages and why lifecycle thinking matters.
AI Agent Monitoring
AI agent monitoring is the practice of tracking agent health, performance, and behavior quality in production. Learn which metrics matter, how monitoring differs from observability, and what to track.
AI Agent Observability
AI agent observability is the practice of capturing and analyzing how AI agents behave in production. Learn the three layers, essential metrics, and why traditional monitoring isn't enough.
AI Agent Rollback
AI agent rollback is the ability to instantly revert a deployed agent to a previous known-good state. Learn how it works, why it matters, and how to implement it.
AI Agent Testing
AI agent testing is the practice of verifying agent behavior through structured evaluations before and during production. Learn the methods, tools, and frameworks for testing non-deterministic AI agents.
AI Agent Traffic Routing
AI agent traffic routing is the practice of directing user requests to specific agent versions based on rules, percentages, or conditions. Learn how it enables canary deployments, A/B testing, and safe rollouts.
AI Agent Versioning
AI agent versioning is the practice of tracking the complete configuration of an AI agent as immutable snapshots. Learn why traditional version control isn't enough and how to implement agent versioning.
Canary Deployment for AI Agents
Canary deployment for AI agents is a release strategy that routes a small percentage of traffic to a new agent version before full rollout. Learn how it works, why it matters, and how to implement it.
Human-in-the-Loop AI
Human-in-the-loop (HITL) AI is a design pattern where AI agents pause and request human approval before taking high-stakes actions. Learn the patterns, trade-offs, and implementation best practices.
LLMOps
LLMOps (Large Language Model Operations) is the practice of managing the lifecycle of LLM-powered applications in production. Learn how it differs from MLOps and AgentOps, and when each applies.