Browse All Entries

Step-by-step implementation guide for building a GraphRAG system.

Comparison of chunking strategies including fixed-size, semantic-aware, and recursive approaches.

Survey of methods for using LLMs to automatically construct knowledge graphs.

AWS's open-source Python library for building graph-enhanced RAG workflows on AWS infrastructure.

Comparison of embedding models for GraphRAG vector search.

Real-world GraphRAG applications across healthcare, legal, finance, manufacturing, and cybersecurity.

Techniques for resolving duplicate entities including fuzzy matching, embedding similarity, and LLM-based resolution.

Comprehensive overview of evaluation metrics covering retrieval, generation, and graph quality.

Five critical optimization techniques that reduced GraphRAG ingestion time from 56 hours to 4.1 hours (13x speedup).

Strategies for optimizing graph database indexing including composite indexes and query caching.

Comprehensive comparison of GraphRAG implementations.

Enterprise customer support using GraphRAG to connect documentation and tickets.

How GraphRAG transforms scientific literature review.

A comprehensive benchmark for evaluating GraphRAG systems across multiple domains.

Strategies for incrementally updating knowledge graphs without full re-indexing.

Comprehensive guide to pgvector index selection, HNSW configuration, and performance optimization.

Recommended PostgreSQL schema patterns for storing documents, embeddings, entities, and relationships.

Optimized prompt templates for high-quality entity and relationship extraction.

Novel approach improving GraphRAG entity extraction through optimized prompt engineering and structured output.

Systematic comparison showing GraphRAG achieves up to 86.31% accuracy on RobustQA benchmarks, a 3x improvement.

Extending RAGAS for GraphRAG-specific metrics.

Strategies for scaling pgvector from small datasets to millions of vectors.

How Text2Cypher enables natural language querying of knowledge graphs.

The recommended production architecture combining PostgreSQL pgvector for semantic search with Neo4j for knowledge graph traversal.

How to implement hierarchical community detection using Neo4j GDS and the Leiden algorithm.

Five distinct architectural patterns for GraphRAG systems, from semantic clustering to query augmentation.

The foundational GraphRAG paper by Microsoft Research introducing the complete pipeline with Leiden clustering and community summaries.

Uses GNNs to perform reasoning over knowledge graphs before passing context to LLMs.

Comprehensive survey categorizing GraphRAG approaches into graph-based indexing, graph-guided retrieval, and graph-enhanced generation.

A lightweight alternative requiring significantly fewer tokens and API calls.

Implementation guide for combining pgvector semantic search with PostgreSQL full-text search using RRF.

Official LangChain integration for Neo4j with graph-based retrievers and Text2Cypher.

LlamaIndex's built-in knowledge graph index with automatic entity extraction.

The official Microsoft GraphRAG implementation with 31.4k+ stars.

A minimal implementation in ~800 lines of Python, ideal for learning.

Comparison of Neo4j deployment options for GraphRAG applications.

Key Graph Data Science algorithms useful for GraphRAG.

Complete guide to building GraphRAG systems with Neo4j.

The official Neo4j GraphRAG Python package for building and querying knowledge graphs with LLMs.

Framework for faithful LLM reasoning with planning-retrieval-reasoning decomposition.

The complete two-phase GraphRAG pipeline covering text chunking, entity extraction, knowledge graph construction, community detection, and multi-strategy querying.

LLMs perform step-by-step reasoning by traversing knowledge graph paths.

Comprehensive roadmap for integrating LLMs with knowledge graphs.

GraphRAG is a structured, hierarchical approach to RAG that uses knowledge graphs instead of naive semantic search, developed by Microsoft Research.