Agentic AI for Platform Engineering: Rethinking developer platforms with agentic AI, RAG, MCP, and governed autonomous operations
暫譯: 平台工程的 Agentic AI：運用 Agentic AI、RAG、MCP 與受治理的自主營運，重新思考開發者平台

Build AI-driven platforms with RAG, MCP, governed agents, and observability, while seeing how Claude Code, Kiro, and Codex CLI fit into the shift from personal coding agents to governed platform agents

Key Features:

- End-to-end coverage from AI-enhanced developer portals to governed agentic operations

- Apply agentic AI patterns for governed remediation, observability, and human-in-the-loop control

- Generate infrastructure-as-code and CI/CD artifacts through conversational AI interfaces

Book Description:

Platform engineers today face a growing challenge: developer platforms must scale faster than teams can maintain them. This book shows how to embed GenAI, RAG, MCP, and governed agents into developer platforms while preserving delivery and governance practices.

The book moves from platform engineering and AIOps foundations to organizational knowledge, conversational interfaces, and agentic operations. Along the way, it places Claude Code, Kiro, and Codex CLI within this evolution, showing how these tools connect to the shift toward governed agents across platforms and organizations. You will build RAG pipelines, connect platform capabilities through MCP, and explore natural-language generation of Terraform and Kubernetes artifacts alongside CI/CD use cases. Hands-on chapters use Backstage, GitOps, Argo CD, Strands, and Langfuse to demonstrate assistants, governed pull-request changes, observability, audit trails, and operational agents.

You will also learn when single- or multi-agent designs are appropriate, evaluate build-versus-buy choices, and address prompt injection, memory poisoning, identity abuse, and compliance. By the end, you will be able to design AI-enhanced developer platforms that improve self-service while keeping high-impact actions observable, auditable, and subject to human approval where risk demands it.

What You Will Learn:

- Build RAG systems grounded in organizational knowledge

- Build governed agents that propose changes through pull requests

- Connect conversational workflows to platform tools with MCP

- Explore where Claude Code and Kiro fit in agentic platform workflows

- Trace agent activity and maintain auditable operational records

- Secure agent workflows with guardrails, identity, and compliance controls

- Compare personal and production agents with governed tool access

- Evaluate build-versus-buy choices for AI platform capabilities

Who this book is for:

This book is for platform engineers, SREs, DevOps professionals, and cloud architects who want to integrate GenAI and agentic AI into internal developer platforms. If you are responsible for building or maintaining internal platforms and want to improve developer self-service, operational intelligence, and governed automation, this book is for you. Technical leaders driving AI adoption within engineering teams will also find valuable strategic guidance. Basic familiarity with cloud computing, infrastructure as code, Kubernetes, CI/CD, and DevOps practices is assumed.

Table of Contents

- Platform Engineering and Developer Experience

- AIOps: A Pre- and Post-Platform Engineering Perspective

- Foundation of Generative AI for Platform Teams

- Driving Platform Intelligence with Organizational Knowledge

- Conversational Interfaces

- Agentic AI and Autonomous Behavior

- Operational Agents in Production

- Security, Compliance, and Governance

- Future Directions