S devlin mira (10 Ergebnisse)

PAP. Zustand: New. New Book. Shipped from UK. Established seller since 2000.

PAP. Zustand: New. New Book. Shipped from UK. Established seller since 2000.

Taschenbuch. Zustand: Neu. Neuware.

Taschenbuch. Zustand: Neu. Neuware - Build AI systems that don't just respond-they collaborate, adapt, and evolve.Large Language Models changed the world.Multi-agent LLM ecosystems will redefine it.In Scaling LLM Agents, Mira S. Devlin takes you inside the next revolution of AI architecture-where networks of specialized LLM agents coordinate, reason, reflect, and self-optimize to accomplish what a single model could never do alone.This is not another 'prompting' book.This is a systems-engineering blueprint for building scalable, resilient, and tool-driven AI ecosystems that behave more like distributed organizations than standalone chatbots.Drawing from modern AI orchestration frameworks, emerging research on distributed cognition, and real-world production patterns, this book shows you how to design agent clusters, shared memory fabrics, routing layers, evaluators, and tool hubs that dynamically adapt and continuously improve.What You'll Learn- How to structure LLM agents into clusters with clear roles, capabilities, and communication protocols- Coordinator and scheduler patterns for robust multi-agent task execution- Decentralized routing fabrics for fast, scalable message passing- Shared vector memory systems for persistent state, grounding, and context fusion- Reflection and optimization loops that help agents correct themselves and learn from outcomes- Tool-driven orchestration using APIs, function calling, and external workflows- Monitoring, evaluation, and telemetry layers to keep your multi-agent system safe, reliable, and transparent- Scalable system topologies for production workloads, real-time reasoning, and enterprise automationWho This Book Is For- AI engineers designing intelligent, high-performance systems- Software architects modernizing applications with agentic patterns- Founders and CTOs building AI-first products- Researchers exploring distributed cognition and emergent behaviors- Developers wanting to go beyond prompts and learn real LLM engineeringIf you've mastered prompting, tinkered with agents, or built early prototypes-this is the book that takes you into the next era: true multi-agent AI ecosystems that scale.Why This Book MattersThe future will not belong to the biggest model.but to the best-organized constellation of collaborating models.This book gives you the architecture, patterns, and practical frameworks to build them.Order your copy today and start building the AI ecosystems of tomorrow.…

Taschenbuch. Zustand: Neu. Neuware.

Taschenbuch. Zustand: Neu. Neuware.

Taschenbuch. Zustand: Neu. Neuware.

Taschenbuch. Zustand: Neu. Neuware.

Taschenbuch. Zustand: Neu. Neuware.

Taschenbuch. Zustand: Neu. Neuware - Amazon Book DescriptionAI can write code faster. But can your engineering process keep up Software development is entering a fundamental transition. AI coding assistants and autonomous development tools can generate code, tests, refactoring plans, documentation, and multi-file implementations at a speed that changes the economics of software production.But faster code generation does not automatically produce better software.Software Engineering in the AI Era examines what happens when implementation becomes dramatically easier-and why code review, testing, architecture, productivity measurement, and developer workflows must evolve with it.Mira S. Devlin presents a practical framework for engineering teams navigating AI-assisted development. Rather than treating AI as either a replacement for developers or simply a faster autocomplete tool, this book examines AI as a change to the entire software engineering system.You will explore how to: - Rethink code review when AI can generate substantial implementations- Build testing strategies that provide evidence rather than simply more test code- Preserve architectural boundaries as implementation becomes cheaper- Identify technical debt and emerging comprehension debt- Evaluate AI-assisted productivity without confusing output with engineering value- Design developer workflows around appropriate human checkpoints- Improve onboarding and knowledge transfer in AI-assisted teams- Understand the risks of generated assumptions, hidden dependencies, and plausible but incorrect implementations- Adapt engineering practices without abandoning proven software engineering fundamentals- Build workflows that balance development speed with correctness, maintainability, and accountabilityThe central idea is simple: When code becomes cheaper to produce, engineering judgment becomes more valuable.AI can generate an implementation. It cannot remove the engineering responsibility to determine whether that implementation is correct, appropriate, understandable, and safe to operate.This book is for software engineers, senior developers, architects, engineering managers, technical leaders, and teams adopting AI-assisted development who want to increase their leverage without allowing development speed to outrun engineering understanding.The future of software engineering is not simply about writing more code. It is about building better systems with greater leverage-and knowing what still requires human judgment. …