Agentic Retrieval-Augmented Generation: Building AI that Searches, Reasons and Acts - Couverture souple

Suman, Vishwanil; Dahiya, Sonia; Sengar, Jitin Kumar

Couverture souple

ISBN 10 : 6630401026 ISBN 13 : 9786630401028

Editeur : LAP Lambert Academic Publishing, 2026

Synopsis

Agentic RAG explores the evolution of Retrieval-Augmented Generation (RAG) into intelligent AI systems that can reason, plan, retrieve knowledge, use external tools, and continuously improve their responses. The book begins with the fundamentals of traditional RAG, its architecture, and its limitations, before introducing Agentic RAG as a more autonomous and adaptive approach. It explains how AI agents understand user intent, rewrite and decompose complex queries, select appropriate knowledge sources, and perform multi-stage retrieval. The book also covers reasoning loops, planning strategies, dynamic tool selection, iterative search, self-correction, and memory management, including short-term, long-term, and semantic memory. Readers will learn how to verify retrieved evidence, reduce hallucinations, implement security guardrails, and evaluate system performance. Finally, the book focuses on building production-ready applications using scalable architectures, multi-agent collaboration, knowledge graphs, and multimodal retrieval.

Détails de l'édition

• Éditeur: LAP Lambert Academic Publishing
• Date d'édition: 2026
• Langue: anglais
• ISBN 10: 6630401026
• ISBN 13: 9786630401028
• Reliure: Tapa blanda
• Nombre de pages: 80
• Coordonnées du fabricant: non disponible
• Personne responsable: non disponible