Agentic Approach to Using Neo4j Knowledge Graphs in Salesforce
summary
Shows a working integration that wires Salesforce LWC + Apex to a Python MCP server acting as a graph RAG pipeline that queries Neo4j and an LLM (Claude) so Salesforce can run natural‑language graph queries. Two demos use Neo4j’s medical sample graph and compare a baseline manual gateway approach versus the pipeline; the vector+cypher hybrid returns noticeably more accurate, grounded answers. Source code, named‑credentials connection, and end‑to‑end results are provided.
AI-written summary · about
chapters (9)
- 0:00Introduction
- 0:48Why integrate Neo4j with Salesforce
- 3:19Use cases for knowledge graphs
- 5:28Agentic approach architecture
- 6:34Traditional method limitations
- 7:01Demo overview and setup
- 8:56Comparison of methods
- 11:34Summary and benefits
- 11:55Q&A
related talks
30mAgentic RAG with Knowledge Graphs Explained
49mConnecting the Dots with Context Graphs
23mConnecting the Dots With Context Graphs
15mGraph-Driven Knowledge Retrieval: Neo4j for Healthcare Building Codes
53mAgents, Tools, and Mcp, Oh My! Next Level AI Concepts for Developers
22mBuilding Cypher Queries With Open-Source LLMs
20mPower Amazon Bedrock applications with Neo4j knowledge graph
08mHarness Graph Power With Agentic AI: An Insights Discovery Use Case