Neo4j MCP Our official Neo4j MCP Server is now available in version 1.0.1 and works with common MCP hosts, such as Claude Desktop, VS Code and others, for all types of Neo4j deployments: Neo4j Aura, Self-Managed, Docker, Neo4j Desktop and Neo4j Sandbox. Get it from Github and explore our MCP Developer pages for more info.
Deploying Single Instance Neo4j On A VPS With Custom Domain Will Lyon provides a solid reference for anyone looking to run a secure, production-ready Neo4j instance outside of managed services. In the article, he deploys Neo4j on a VPS, covering everything from Java installation and APOC setup to domain configuration, HTTPS via Let’s Encrypt, firewall rules, and restoring from backups.
Introducing Neo4j’s Native Vector Data Type
David Pond gives us the latest about the new built-in VECTOR data type. Embedding vectors can be stored as fixed-length, typed properties rather than generic lists. This change simplifies application code, enforces vector shape/dtype constraints for safety, and unlocks future vector-specific optimisations in storage, indexing and GenAI workflows.
Constructing Knowledge Graphs with Neo4j GraphRAG for Python
This newest course from GraphAcademy shows you how to use the official neo4j-graphrag Python package to build knowledge graphs from unstructured text, enrich them with structured data, and create retrievers that let LLMs query graphs via Cypher or hybrid vector-graph search.
|
