Agentic AI in Action: Architecting Multi-Agent Systems, Autonomous Workflows, and Goal-Oriented Enterprise Systems

Name: Agentic AI in Action: Architecting Multi-Agent Systems, Autonomous Workflows, and Goal-Oriented Enterprise Systems
SKU: 25aa9ae5fff9bf543a196e72ebd4bd3d
Price: 24.99 EUR
Availability: InStock

1/1

Afbeelding van Agentic AI in Action: Architecting Multi-Agent Systems, Autonomous Workflows, and Goal-Oriented Enterprise Systems

Bol

Prijzen vanaf

24,99

Uitgelicht

	24,99	Naar shop
VERGELIJK ALLE AANBIEDERS (1)

Beschrijving

Bol Agentic AI in Action: Architecting Multi-Agent Systems, Autonomous Workflows, and Goal-Oriented Enterprise SystemsThe era of stateless request-response architectures and basic RAG pipelines is ending. The future of enterprise software belongs to autonomous, goal-directed agentic systems capable of reasoning, planning, and executing complex workflows without continuous human intervention.In Agentic AI in Action, Diane B. Cray delivers the definitive engineering blueprint for bridging the gap between probabilistic generative AI and resilient, production-grade machine learning pipelines. This is not a guide to basic prompt engineering. It is a rigorous architectural manual designed for senior software engineers, system architects, and MLOps professionals tasked with building secure, scalable, and fault-tolerant AI systems.Moving systematically from the cognitive anatomy of a single agent to the orchestration of distributed swarms, this book provides the theoretical foundations and targeted Python implementations required to deploy autonomous systems in hostile enterprise environments.Inside the Book: - Cognitive Architectures & Reasoning: Move beyond linear chains. Implement ReAct (Reason + Act) loops, Plan-and-Solve task decomposition, Tree-of-Thought exploratory search, and runtime reflexion mechanisms for self-correcting agents.- Multi-Agent Swarm Design: Orchestrate decentralized peer-to-peer swarms and hierarchical command structures. Synchronize shared state using Blackboard architectures and handle inter-agent communication via enterprise message brokers like Kafka and RabbitMQ.- Conflict Resolution & Consensus: Manage divergent outputs through weighted majority voting, quorum logic, and adversarial red-teaming critique loops.- Production-Grade Type Safety: Enforce strict JSON schemas and Pydantic models to translate probabilistic LLM semantic intent into deterministic, type-safe API calls. Implement dynamic capability routing and exponential backoff circuit breakers.- State Management & Resilience: Engineer Directed Acyclic Graphs (DAGs) for workflow control, distributed checkpointing for multi-day tasks, transactional Saga rollbacks, and time-travel debugging for non-deterministic executions.- Zero-Trust Security & Sandboxing: Protect enterprise infrastructure by neutralizing indirect prompt injections and data exfiltration. Deploy containerized sandboxes (Docker, gVisor, WebAssembly) for untrusted code execution and enforce strict Role-Based Access Control (RBAC) and PII redaction.- Observability & Scale: Solve the observability crisis in non-linear systems. Implement distributed tracing with OpenTelemetry, enforce real-time token budget governance, and automate continuous improvement pipelines using LLM-as-a-Judge and agentic chaos engineering.Stop building reactive copilots. Master the integration of semantic reasoning with deterministic execution, and architect the autonomous enterprise.

Lees meer

Vergelijk aanbieders (1)

Sorteren op:

€ 24,99 € 2,99 verzendkosten Totaal € 27,98

Beschrijving (1)

Agentic AI in Action: Architecting Multi-Agent Systems, Autonomous Workflows, and Goal-Oriented Enterprise SystemsThe era of stateless request-response architectures and basic RAG pipelines is ending. The future of enterprise software belongs to autonomous, goal-directed agentic systems capable of reasoning, planning, and executing complex workflows without continuous human intervention.In Agentic AI in Action, Diane B. Cray delivers the definitive engineering blueprint for bridging the gap between probabilistic generative AI and resilient, production-grade machine learning pipelines. This is not a guide to basic prompt engineering. It is a rigorous architectural manual designed for senior software engineers, system architects, and MLOps professionals tasked with building secure, scalable, and fault-tolerant AI systems.Moving systematically from the cognitive anatomy of a single agent to the orchestration of distributed swarms, this book provides the theoretical foundations and targeted Python implementations required to deploy autonomous systems in hostile enterprise environments.Inside the Book: - Cognitive Architectures & Reasoning: Move beyond linear chains. Implement ReAct (Reason + Act) loops, Plan-and-Solve task decomposition, Tree-of-Thought exploratory search, and runtime reflexion mechanisms for self-correcting agents.- Multi-Agent Swarm Design: Orchestrate decentralized peer-to-peer swarms and hierarchical command structures. Synchronize shared state using Blackboard architectures and handle inter-agent communication via enterprise message brokers like Kafka and RabbitMQ.- Conflict Resolution & Consensus: Manage divergent outputs through weighted majority voting, quorum logic, and adversarial red-teaming critique loops.- Production-Grade Type Safety: Enforce strict JSON schemas and Pydantic models to translate probabilistic LLM semantic intent into deterministic, type-safe API calls. Implement dynamic capability routing and exponential backoff circuit breakers.- State Management & Resilience: Engineer Directed Acyclic Graphs (DAGs) for workflow control, distributed checkpointing for multi-day tasks, transactional Saga rollbacks, and time-travel debugging for non-deterministic executions.- Zero-Trust Security & Sandboxing: Protect enterprise infrastructure by neutralizing indirect prompt injections and data exfiltration. Deploy containerized sandboxes (Docker, gVisor, WebAssembly) for untrusted code execution and enforce strict Role-Based Access Control (RBAC) and PII redaction.- Observability & Scale: Solve the observability crisis in non-linear systems. Implement distributed tracing with OpenTelemetry, enforce real-time token budget governance, and automate continuous improvement pipelines using LLM-as-a-Judge and agentic chaos engineering.Stop building reactive copilots. Master the integration of semantic reasoning with deterministic execution, and architect the autonomous enterprise.

Lees meer