OpenClaw + n8n + Neo4j: Build a YouTube Knowledge Graph with Gemini Embeddings

Build a YouTube video knowledge graph with OpenClaw, n8n, Gemini embeddings, and Neo4j.

In this tutorial, we create a practical AI automation workflow that connects OpenClaw, n8n, the YouTube Data API, Gemini embeddings, and Neo4j. From Telegram, OpenClaw triggers an n8n workflow, collects official YouTube metadata, generates Gemini embeddings, and stores the result in Neo4j as part of a video knowledge graph.

This tutorial builds on previous OpenClaw and n8n workflow videos. In this episode, we extend the pipeline by adding Gemini embeddings and Neo4j graph database storage. We also create a new OpenClaw skill so the AI agent can trigger the full workflow from Telegram.

📝 Resources:
All scripts and prompts used in this tutorial will be available on the LBSocial blog (URL coming soon!). https://www.lbsocial.net

What you will learn:
- Create a free Neo4j database instance
- Create and store a Gemini API key in n8n
- Install the n8n Neo4j community node
- Duplicate and extend an existing YouTube metadata workflow
- Build a document_text field for embeddings
- Call the Gemini embedding API from n8n
- Merge embeddings with YouTube metadata
- Generate a Cypher query with JavaScript
- Store videos, channels, topics, and embeddings in Neo4j
- Create an OpenClaw skill to trigger the workflow from Telegram
- Verify the video knowledge graph in Neo4j

Timeline:
00:00 Demo: OpenClaw processes a YouTube video
00:43 View the video knowledge graph in Neo4j
01:34 Tutorial overview
02:42 Create Neo4j and Gemini credentials
04:29 Install the Neo4j node and configure n8n credentials
06:11 Duplicate and extend the n8n workflow
07:17 Build document_text and generate Gemini embeddings
09:26 Merge embeddings and write data to Neo4j
11:13 Verify the graph and embeddings in Neo4j
12:13 Reconnect the webhook and publish the workflow
13:05 Create and test the OpenClaw skill
14:54 Closing summary

Related playlists:

https://www.youtube.com/playlist?list=PLHutrxqbP1BxtprL8ALskhFTalhG9u-eF
https://www.youtube.com/playlist?list=PLHutrxqbP1BwAQf6dROCLZqK0PLpGu35J
https://www.youtube.com/playlist?list=PLHutrxqbP1ByyhtEsXvW7ZU5i6FyxTMjr

Tools used:
OpenClaw
n8n
Neo4j AuraDB
Google AI Studio
Gemini Embeddings
YouTube Data API
Telegram

This workflow is a foundation for future GraphRAG tutorials, where OpenClaw can query the Neo4j graph, perform semantic search, retrieve related videos and topics, and generate answers based on your own video knowledge base.

#OpenClaw #n8n #Neo4j #Gemini #GeminiEmbeddings #YouTubeAPI #GraphRAG #AIAgents #KnowledgeGraph #WorkflowAutomation #LBSocial