Skip to content

description: The why and what: what agents and AgentOps are, the open-source ecosystem, and model providers.

0. Overview

What will you learn in this chapter?

This chapter explains when an agent is justified, maps the AgentOps lifecycle, assigns ownership across the open-source stack, separates the required OSS path from optional hosted services, and makes local Qwen3 the default provider choice.

This chapter covers:

  • 0.0. Course: what, who, and why — prerequisites, time, cost, and how to use this course.
  • 0.1. Agents: what an AI agent is, the agentic loop, common patterns, and when not to use one.
  • 0.2. AgentOps: the AgentOps lifecycle and how MLOps, LLMOps, and AgentOps relate.
  • 0.3. Ecosystem: Ownership boundaries across ADK, agentgateway, kagent, MLflow, OTel, MCP, A2A, AAIF, and CNCF.
  • 0.4. Providers: Start with local Qwen3, then compare optional Gemini or Vertex/GKE explicitly.
  • 0.5. Resources: Primary documentation, OSS development tools, community routes, and a glossary.
  • 0.6. Troubleshooting: Symptom-first fixes for the most common setup and runtime failures.