Best Podcasts and Videos About Graph Engineering

Best for: Practitioner trade-offs and observability

Pydantic founder Samuel Colvin explains why he moved from a single “god model” toward compound graphs, testing and observable agent runs.

Best for: Shared state and governance

A production-minded discussion of state, graph semantics and governance—the coordination layer that generic multi-agent explainers usually omit.

Best for: A precise Simplified Chinese explanation

Defines an executable work graph of agents, functions, tools and human checkpoints, then covers recovery, observability and when not to use one.

Best for: The current debate in context

A timely audio account of the new label and its link to agents, tools, knowledge and people, without treating every multi-agent system as a graph.

Best for: State, persistence and human-in-the-loop

LangGraph maintainer Sydney Runkle walks through stateful execution, durable resumability, memory and human intervention in a real Python framework.

Best for: A German workflow-versus-agent critique

INNOQ engineers separate deterministic workflows from autonomous agents, then examine LangGraph, long-running execution, retries and the final 20% of production work.

Best for: Designing auditable evaluation gates

Laurent Zhang separates deterministic scorers from LLM judges and connects evaluation, orchestration and production—the checkpoints that make an agent graph measurable.

Best for: Nodes, conditional edges and state transitions

A developer-to-developer LangGraphJS walkthrough covers nodes, tools, conditional edges, state transitions, context and error handling.

Best for: The framework’s original graph model

LangChain’s official walkthrough treats each agent as a node, connections as edges and graph state as the communication surface across three patterns.

Best for: A live graph from whiteboard to terminal

Draws the AI engineering layers first, then runs two Waku Agent graphs so loops and graph orchestration can be compared in a real tool.

Best for: Comparing swarm and supervisor topologies

LangChain’s official demo compares routing patterns, then shows handoffs, shared state, checkpointers and thread continuity in Studio.

Best for: Designing breakpoints and human feedback

An official demo pauses the graph at a breakpoint, gathers user clarification or feedback and resumes from persisted state.

Best for: Building conditional graphs in Python

Builds simple and complex agents, adds structured classification and conditional edges, then renders the graph for inspection.

Best for: A technical Simplified Chinese overview

Explains nodes, edges, state and policy, then compares diamond, supervisor and pipeline topologies with independent verification.

Best for: A Spanish software-engineering explanation

A developer-focused explanation from a Spanish computer-science educator, useful for learning the architecture without translating English terminology.

Best for: A Portuguese business workflow demo

Shows a real operational agent graph on screen and explains when explicit orchestration pays off—and when a simpler setup is enough.