Summary

Keywords

neo4j graph databases graphs LLM retrieval augmented generation RAG vector database vector search GenAI AI Artificial intelligence nodes25

Full Transcript

In this session, Linghua Jin, CEO of CocoIndex, will demonstrate how to build a real-time knowledge graph from unstructured documents using LLMs and graph databases. As the volume of content grows, extracting and maintaining structured insights across evolving documentation has become a significant challenge. This talk introduces CocoIndex, an open-source framework designed to automate this process by continuously processing markdown files, extracting semantic relationships with GPT-4o, and exporting them to Neo4j in real time. You will learn how to ... - use LLMs to identify entity relationships and mentions within technical documents. - structure these insights as subject–predicate–object triples. - stream this data into a live Neo4j graph using CocoIndex's declarative pipeline. - visualize and query the resulting knowledge graph for insights and traceability. Whether you're building developer portals, internal search systems, or AI agents with contextual memory, this session will equip you with a practical and scalable blueprint for transforming text into knowledge. Speaker: Linghua Jin View Presentation: https://drive.google.com/file/d/15WM6CXAFyw5JiBYGwk-zyNZCeQJ71EDQ/view?usp=drive_link Resources: Get Started with Aura - https://bit.ly/3LOLrjh Deployment Center - https://bit.ly/4jOelM3 Ground AI Systems and Agents with Neo4j - https://bit.ly/4oVsnyb #nodes2025 #neo4j #graphdatabase #graphrag #knowledgegraph