Graph Engineering for Agentic AI Systems. Este artículo no está disponible.Graph Engineering for Agentic AI Systems

Design the topology. Don't inherit it.

Most agent systems get their shape by accident — a prompt, then a tool, then a retry, then a second agent because the first one got long. Nobody writes the topology down.

Graph engineering is the discipline of declaring it on purpose: the nodes that should be designed, the edges that are legal, the state that travels with each run, and the gates that sit on every irreversible step.

Ken Huang writes for engineers who have to ship these systems and then operate them. Fourteen chapters cover:

• Execution graphs and memory graphs — and what conflating them costs you
• What actually runs under LangGraph, AutoGen GraphFlow, and ADK's Workflow Runtime
• When a loop is the right answer and a graph is over-engineering
• Knowledge graphs as durable agent memory
• Security, identity, and governance for agent graphs
• Testing, evaluation, and verification that survive production
• The computer science underneath — and what stays true after the name fades

This paperback edition runs 364 pages and includes a full back-of-book index. Companion code and homework live on GitHub with links in the book.

About the author. Ken Huang is CEO and Chief AI Officer at DistributedApps.ai, Co-Chair of AI Safety groups at the Cloud Security Alliance and the OWASP AIVSS project, and Adjunct Professor at the University of San Francisco. He coauthored OWASP's Top 10 for LLM Applications, contributes to the NIST Generative AI Public Working Group, and has published with Springer, Cambridge, Packt, China Machine Press, IEEE, ACM, and Wiley.