Skip to content

Enterprise Agent Platform Engineering

This book is for teams building enterprise-grade Agent platforms. It covers the engineering system behind production Agents: data foundations, model inference, knowledge engineering, Agent Runtime, tool ecosystems, evaluation, deployment, frontend interaction, security, compliance, and organizational governance.

The English edition follows the same structure as the Chinese edition and focuses on stable technical chapters and case-review methods backed by explicit evidence standards.

Edition Entry Points

Edition Entry
Chinese Available from the site language switcher
English Start with Part I

Front Matter

Page Purpose
Abbreviations Defines common abbreviations and the terminology used in this book
Preface Explains the platform-engineering viewpoint and release boundary
Acknowledgements Acknowledgements page
Front Matter Guide Gives reading paths by role and problem type
Contributors Contributors page

Quick Navigation

Part Topic
Part I Overview and Platform Perspective Agent fundamentals, platform boundaries, AI-native business systems, and the full-book map
Part II Models and Inference Model selection, local inference, inference optimization, structured output, and customization
Part III Data Infrastructure Ingestion, lakehouse, OLAP, streaming, orchestration, quality, metadata, and metrics
Part IV Vectors, Retrieval, and Knowledge Engineering Embeddings, reranking, vector databases, document parsing, RAG, and knowledge engineering
Part V Agent Capabilities Runtime, Tool Registry, MCP, Planner, Workflow, Memory, multi-agent systems, and protocols
Part VI DataAgent Deep Dive Semantic layer, NL2SQL, Python analysis, visualization, reporting, and ecosystem comparison
Part VII Observability, Evaluation, and Cost Trace, offline evaluation, online evaluation, cost governance, and SLOs
Part VIII Deployment and Infrastructure GPU scheduling, model deployment, LLM Gateway, GitOps, and edge inference
Part IX Frontend, Interaction, and Multimodality Conversational UI, Generative UI, multimodal input, and voice Agents
Part X Security, Compliance, and Organization Security, Guardrails, regulation, organization, and platform evolution
Part XI Case Methodology Case admission, review, and platform consolidation standards

Reading Paths

Role Recommended Path
AI platform leader / CTO Part I -> Part V -> Part VI -> Part X
Architect Part I -> Part II -> Part III -> Part IV -> Part V -> Part VIII
Data intelligence engineer Part III -> Part IV -> Part VI -> Part VII
AI application developer Part II -> Part V -> Part IX -> mini-platform
Security / compliance owner Part I -> Part VII -> Part X

Local Validation

bash scripts/check_all.sh
python -m mkdocs build --strict --clean --site-dir /tmp/enterprise-agent-book-site

The first command checks chapter structure, terminology, sensitive information, and mini-platform tests. The second command verifies that the web-book project can build in strict mode. Together, they cover the main local quality gates before a submission.