GraphRAG
GraphRAG is a retrieval-augmented generation approach that uses graph-structured knowledge and relationships to retrieve context for language-model responses.
Knowledge Graphs and LLMs in Action
GraphRAG is a retrieval-augmented generation approach that uses graph-structured knowledge to ground language-model outputs.
- How it is used
- Grounds language-model question answering and agents in structured, current, traceable knowledge.
- How deep it goes
- Explains graph design, retrieval paths, agent orchestration, metadata, provenance, and implementation tooling rather than only defining the technique.
- Examples
- Uses a Lauritsen–cyclotron retrieval failure, Rockefeller Archive Center questions, graph retrievers, and a Neo4j implementation listing.
- What you can do after
- Operate GraphRAG agents with KG retrievers and KG-enhanced document retrievers to produce more precise, grounded answers.
Read how Knowledge Graphs and LLMs in Action explains GraphRAG →
The Art of AI Product Development
A retrieval-augmented generation approach that uses graph-structured knowledge and entity relationships to retrieve contextual information.
- How it is used
- Improves retrieval for technical, legal, research-heavy, and complex multistep applications.
- How deep it goes
- Explains graph construction, traversal, pipeline integration, and domain tailoring without implementation code.
- Examples
- Uses finance relationships among regulations, companies, and market events as the domain-modeling example.
- What you can do after
- Apply it via a RAG pipeline that retrieves connected, context-rich evidence for complex queries.
Read how The Art of AI Product Development explains GraphRAG →
Essential GraphRAG
GraphRAG is a retrieval-augmented generation approach that incorporates graph-structured knowledge into information retrieval and response generation.
- How it is used
- Supports context-rich, interconnected retrieval for complex queries across diverse document collections.
- How deep it goes
- Builds an end-to-end workflow from entity extraction through graph indexing, community summarization, and global or local retrieval.
- Examples
- Uses The Odyssey, recurring characters, gods, mystical weapons, and cross-chunk entity summaries as evaluation artifacts.
- What you can do after
- Build it from scratch by modeling entities, constructing graphs, implementing retrieval workflows, and evaluating system quality.
Domain-Specific Small Language Models
A retrieval-augmented generation approach that uses knowledge graphs and community summaries to support multihop reasoning over document collections.
- How it is used
- Supports multihop question answering over document collections through graph-based retrieval and community-level summarization.
- How deep it goes
- Explains global and local query paths, indexing persistence, and deployment constraints without detailing graph-construction internals.
- Examples
- Uses CLI installation, workspace initialization, indexing, and global or local querying as the operational workflow.
- What you can do after
- Operate GraphRAG through its Python package and CLI, selecting global or local search and configuring supported persistence services.
Read how Domain-Specific Small Language Models explains GraphRAG →
AI Agents and Applications
GraphRAG is a retrieval-augmented generation pattern that uses graph-structured knowledge for retrieval, reasoning, and response generation.
- How it is used
- Uses structured knowledge graphs to retrieve connected facts, support complex reasoning, and generate natural-language answers.
- How deep it goes
- Explains graph construction, query generation, execution, and response synthesis, but defers detailed KG-RAG implementation.
- Examples
- Uses Roberto, Inter Milan, and Serie A as graph entities connected by typed relationships and triples.
- What you can do after
- Build a GraphRAG workflow that generates Cypher or SPARQL queries and converts retrieved graph data into answers.
Related concepts
- knowledge graphuses · A knowledge graph is a graph-structured representation of entities and their semantically meaningful relationships, often including associated attributes or facts.
- Leiden algorithmuses · The Leiden algorithm is a graph community-detection method that iteratively refines and aggregates partitions into well-connected communities while optimizing a quality function.
- retrieval-augmented generationkind of · Retrieval-augmented generation is a technique that retrieves relevant information from external sources and supplies it as context to a generative model when producing an output.
These guides are generated from the text of Manning books by Manning's knowledge graph and link to the sections they draw on. The full books are available on liveBook and at manning.com.