GraphRAG Chatbot for Specialized Knowledge Retrieval - n8n Workflow

• Knowledge Managers and Researchers needing fast, semantic retrieval.


• Developers seeking simple RAG solutions without maintaining a vector database.


• Users wanting ready-to-use n8n templates for conversational AI.


• Businesses building expert support bots based on proprietary data graphs.

This n8n workflow streamlines the process of building an AI-powered expert chatbot. Unlike traditional Retrieval-Augmented Generation (RAG) pipelines that require complex setups involving vector stores, this solution leverages the specialized InfraNodus GraphRAG n8n node. This node connects directly to a conceptual graph knowledge base to generate highly contextual and relevant answers based on structured data and custom ontologies.

This simple, three-step n8n template is ideal for testing internal knowledge bases quickly via the built-in n8n chat functionality, or it can be easily adapted using a standard Webhook n8n trigger for external deployment via the n8n Chat Widget. This approach provides a powerful and lightweight alternative for creating intelligent Q&A interfaces.

The operation of this n8n workflow is concise and highly focused on specialized data retrieval:

1. Initiation (The n8n trigger): The process begins with the When chat message received n8n trigger. This node captures the user's input query (chatInput) in a conversational setting, often used for testing within the n8n environment itself. Alternatively, a standard Webhook n8n trigger can be used for external application integration.


2. Knowledge Query: The captured user prompt is passed to the specialized Get a response from knowledge base n8n node (InfraNodus GraphRAG). This node uses the prompt to query the defined knowledge graph, utilizing GraphRAG technology to synthesize an expert answer. This step is the core of the n8n workflow, fetching the necessary context.


3. Response Generation: The synthesized answer, extracted from the output path {{ $json.aiAdvice[0].text }}, is sent to the Respond to Chat n8n node. This node immediately delivers the knowledge base's response back to the user, completing the conversational loop initiated by the n8n trigger.

To deploy this expert n8n workflow, follow these steps:

1. Import the n8n template: Copy the provided JSON code and import it directly into your n8n instance.


2. InfraNodus Credentials: You must configure the InfraNodus API credential.

1. Configure GraphRAG Node: Update the Get a response from knowledge base n8n node. Ensure you specify the correct name parameter corresponding to the graph (knowledge base) you wish to query in InfraNodus.


2. Activate: Save the n8n workflow and click 'Activate'. You can now test the n8n trigger using the built-in chat interface.

When chat message received (n8n trigger)
Function: Acts as the conversational entry point for the n8n workflow, capturing the user's query.
Key Configuration: Set up to capture chatInput which dynamically feeds the query into the next node.
Get a response from knowledge base (InfraNodus n8n node)
Function: The central processing step. This n8n node utilizes GraphRAG to retrieve and synthesize an expert response based on the defined knowledge graph in InfraNodus.
Key Configuration: Requires InfraNodus API credentials and the prompt parameter set to ={{ $json.chatInput }} to pull the query from the initial n8n trigger.
Respond to Chat
Function: Sends the final, synthesized answer back to the user interface, completing the interaction loop created by the initial n8n trigger.
Key Configuration: The message parameter is dynamically set using the output of the GraphRAG node: ={{ $json.aiAdvice[0].text }}.
Webhook / Respond to Webhook (Disabled)
* Function: Shows the alternative path for exposing this n8n workflow publicly as a standard API endpoint, allowing deployment outside of the internal n8n chat environment.