Agentic RAG

Concept

An evolution of retrieval-augmented generation in which the language model plans, chooses data sources, evaluates retrieved results and queries again in a loop instead of answering from a single retrieval pass.

Title graphic: Agentic RAG (Concept)

Type	Concept

Key points about Agentic RAG

Agentic RAG on Azure turns Copilot from a context tourist into a reasoning engine. · Build Traceable Copilots with Agentic RAG on…
In Agentic RAG the model takes part in deciding what information is needed and how to obtain it, following the ReAct pattern of Reason, Act, Reason, Act. · The Copilot Assumption That's Holding You Back…

Episodes about Agentic RAG

RAG on Azure — Simply Explained

2026-07-16 · Video on YouTube

Retrieval-Augmented Generation (RAG) is one of the most important techniques behind modern AI applications, enabling large language models to answer questions using your own business data instead of relying only on their training knowledge. In this episode of…

Build Traceable Copilots with Agentic RAG on Azure

2025-11-17 · Video on YouTube

Your Copilot isn’t smart – it’s a very expensive autocomplete. In this episode, we break down why classic RAG (retrieve-augment-generate) quietly fails the moment your truth lives in more than one system, and how “agentic RAG” on Azure turns Copilot from a…

The Copilot Assumption That's Holding You Back — Agentic…

2026-08-11 · Video on YouTube

Many teams still treat Copilot as a search box and blame the tool when answers come back incomplete. This episode explains agentic RAG, the orchestration loop behind modern Copilot, for IT decision makers and Microsoft 365 architects. You learn when agentic…

Timeline of episodes about Agentic RAG

Articles about Agentic RAG (2)

Agentic AI in Action: Transforming Paperwork into Voice-First Workflows
2026-08-13 · Mirko Peters
For years, the enterprise software conversation has been dominated by the desk-bound knowledge worker. We talk endlessly about email management, document drafting, …
Cost and Latency Trade-Offs in Agentic RAG vs Traditional Retrieval
2026-08-11 · Mirko Peters
Discover why Agentic RAG is 2-10x more expensive and slower than traditional RAG, and learn how to choose the right architecture for your enterprise AI.

Frequently asked questions

What is agentic RAG?

It is RAG in which Planner, Retriever and Verifier agents work together, query several systems and cross-check answers. This goes beyond a single retrieve-augment-generate pass and produces evidence-linked insights. · From the episode: Build Traceable Copilots with Agentic RAG on…

How do agentic RAG agents divide the work?

The Planner decides the approach, the Retriever fetches from Fabric, SharePoint and external data, and the Verifier checks the results. Running on Azure AI Agent Service, they yield traceable answers. · From the episode: Build Traceable Copilots with Agentic RAG on…

How does agentic RAG respect permissions?

It uses On-Behalf-Of auth, row- and column-level security and Purview labels. The AI then sees only what the human user may see, so answers follow the same rules as the person asking. · From the episode: Build Traceable Copilots with Agentic RAG on…

Glossary

Community wiki

Short info above, community knowledge below: add practical tips, examples or corrections. New contributions are reviewed first; trusted members publish directly.

Experts and companies for this topic

Nobody has listed this topic yet. Add it to your profile.

Sources

Feed for this topic (RSS): entity/agentic-rag/feed.xml
M365.FM episodes on this topic: 3
Articles on this topic: 2

Maintained by Mirko Peters (Microsoft MVP) · Last updated: 2026-11-03