Agentic AI Simplified, Vol. 2 is a hands-on guide for going from toy AI agents to production-ready multi-agent systems on AWS, Azure, and Google Cloud.
It covers six agent frameworks (LangGraph, CrewAI, AutoGen, OpenAI Agents SDK, Google ADK, Claude SDK) with trade-off comparisons, production-grade RAG (traditional, agentic, adaptive) with pgvector and hallucination control, all three major cloud AI platforms, a full-stack deployment (FastAPI, pgvector, Azure Container Apps, Terraform, JWT auth), AI governance (RAGAS evaluation, human-in-the-loop, observability, audit trails), a full enterprise case study (insurance claims processing with blockchain and multi-agent orchestration), and emerging patterns like voice, image agents, and MCP.
It's aimed at software engineers, ML practitioners, and solution architects with Python basics who want to move from prototyping to shipping reliable, auditable systems. Works standalone, though Vol. 1 readers get a smoother on-ramp.
The pitch: most books explain the "what" of agents; this one focuses on the "how" — framework choice, debugging, compliance, real evaluation — with a full production curriculum rather than disconnected tutorials.