Agentic Workflows with Python and LangGraph

Reactive PublishingTake your software architecture beyond single-prompt execution. Agentic Workflows with Python and LangGraph provides software engineers, data scientists, and system architects with a clear blueprint for designing, orchestrating, and maintaining reliable multi-agent AI environments.As large language models transition from simple chat assistants to active problem solvers, software systems require robust control flow, state persistence, and memory handling. This book covers the core principles needed to construct deterministic, stateful graph architectures using Python and LangGraph-allowing autonomous agents to collaborate, self-correct, and execute complex business logic seamlessly.What You Will Learn:Core Architectural Foundations: Transition from linear prompt chains to cyclical, graph-based workflows that manage state across long-running tasks.Multi-Agent Coordination: Design patterns for agent orchestration, including supervisor routes, hierarchical teams, and peer-to-peer collaboration models.State & Memory Management: Implement short-term working memory, long-term semantic storage, and transactional state checkpoints for reliable execution.Human-in-the-Loop Integration: Build fail-safes, approval checkpoints, and manual intervention loops directly into autonomous workflows.Tool Execution & Routing: Enable safe, dynamic tool selection so agents can execute external functions, query databases, and parse structured output.Whether you are looking to modernise existing automation pipelines or construct new autonomous platforms from the ground up, this practical guide offers the precise technical depth needed to turn non-deterministic models into predictable enterprise infrastructure.

Reactive PublishingTake your software architecture beyond single-prompt execution. Agentic Workflows with Python and LangGraph provides software engineers, data scientists, and system architects with a clear blueprint for designing, orchestrating, and maintaining reliable multi-agent AI environments.As large language models transition from simple chat assistants to active problem solvers, software systems require robust control flow, state persistence, and memory handling. This book covers the core principles needed to construct deterministic, stateful graph architectures using Python and LangGraph-allowing autonomous agents to collaborate, self-correct, and execute complex business logic seamlessly.What You Will Learn:Core Architectural Foundations: Transition from linear prompt chains to cyclical, graph-based workflows that manage state across long-running tasks.Multi-Agent Coordination: Design patterns for agent orchestration, including supervisor routes, hierarchical teams, and peer-to-peer collaboration models.State & Memory Management: Implement short-term working memory, long-term semantic storage, and transactional state checkpoints for reliable execution.Human-in-the-Loop Integration: Build fail-safes, approval checkpoints, and manual intervention loops directly into autonomous workflows.Tool Execution & Routing: Enable safe, dynamic tool selection so agents can execute external functions, query databases, and parse structured output.Whether you are looking to modernise existing automation pipelines or construct new autonomous platforms from the ground up, this practical guide offers the precise technical depth needed to turn non-deterministic models into predictable enterprise infrastructure.