Agentic Retrieval-Augmented Generation : A Survey On Agentic RAG

Overview of Agentic RAG

Recent Update (2025-02-04):

Check section 4 in the table of contents in this repo for the new Agentic Workflow Patterns. New images have been added to enhance the Overview of Agentic RAG. The paper is also updated.

Abstract

Agentic Retrieval-Augmented Generation ( Agentic RAG) represents a transformative leap in artificial intelligence by embedding autonomous agents into the RAG pipeline. This repository complements the survey paper "Agentic Retrieval-Augmented Generation (Agentic RAG): A Survey On Agentic RAG," providing insights into:

• Foundational principles, including Agentic Patterns such as reflection, planning, tool use, and multi-agent collaboration.
• A detailed taxonomy of Agentic RAG systems, showcasing frameworks like single-agent, multi-agent, hierarchical, corrective, adaptive, and graph-based RAG.
• Comparative analysis of traditional RAG, Agentic RAG, and Agentic Document Workflows (ADW) to highlight their strengths, weaknesses, and best-fit scenarios.
• Real-world applications across industries like healthcare, education, finance, and legal analysis.
• Challenges and future directions in scaling, ethical AI, multimodal integration, and human-agent collaboration.

This repository serves as a comprehensive resource for researchers and practitioners to explore, implement, and advance the capabilities of Agentic RAG systems.

Table of Contents

1. 1.📜 Abstract
2. 2.🧩 Introduction
3. 3.🤖 Agentic Patterns
4. 4.🔄 Agentic Workflow Patterns
5. 5.🛠️ Taxonomy of Agentic RAG Systems
6. 6.🔍 Comparative Analysis of Agentic RAG Frameworks
7. 7.💼 Applications
8. 8.🚧 Challenges and Future Directions
9. 9.🛠️ Implementation of RAG Agentic Taxonomy: Techniques and Tools
10. 10.📰 Blogs and Tutorials on Agentic RAG
11. 11.🖊️ Noteworthy Related Concepts
12. 12.💡 Practical Implementations and Use Cases of Agentic RAG
13. 13.📚 References
14. 14.🖊️ How to Cite

Introduction

Retrieval-Augmented Generation (RAG) systems combine the capabilities of large language models (LLMs) with retrieval mechanisms to generate contextually relevant and accurate responses. While traditional RAG systems excel in knowledge retrieval and generation, they often fall short in handling dynamic, multi-step reasoning tasks, adaptability, and orchestration for complex workflows.