GraphRAG-Powered AI Agents: Transforming Corporate Metadata into Intelligent Knowledge
summary
A system combines Neo4j knowledge graph with LLM-based agents and pipeline tooling to turn catalog metadata and ingested PDFs into a searchable knowledge base. It links tables, column descriptions, Bill of Materials and external docs, runs graph algorithms for centrality and similarity, and produces ranked lists (e.g., supplier audit candidates) and clickable reports that point to Snowflake tables. A demo using a cloud desktop black box and MCP server integration successfully finds lineage paths, surfaces EU regulation‑related tables, and auto‑generates dashboardable summaries.
AI-written summary · about
chapters (9)
- 0:00Introduction and speaker intro
- 0:25Speaker background and company overview
- 1:07What is metadata and its types
- 3:38Why metadata matters for enterprises
- 5:46GraphRAG and knowledge graphs explained
- 7:54Building the knowledge base with unstructured data
- 8:57Agent framework and security layers
- 10:44MCP servers integration demo
- 13:32Conclusion and future applications
related talks
20mPractical GraphRAG: Making LLMs smarter with Knowledge Graphs
28mData-Empowered AI Assistants
12mAgentic GraphRAG: Autonomous Knowledge Graph Construction and Adaptive Retrieval
08mHarness Graph Power With Agentic AI: An Insights Discovery Use Case
32mGraphRAG: Powering Up LLMs with Knowledge Graphs
21mAgentic GraphRAG: Autonomous Knowledge Graph Construction and Adaptive Retrieval
21mFrom Data to Knowledge to Action - The Graph Intelligence Platform
39mEnhancing RAG with Neo4j Knowledge Graph