n8n AI Chatbot with Infranodus Knowledge Graph for Enhanced Responses

This n8n workflow provides a framework for building an AI chatbot that leverages an Infranodus knowledge graph to enhance its responses. It uses an n8n Form Trigger to initiate interactions, allowing users to submit queries, and an HTTP Request node to interact with external services, likely including an AI model and the Infranodus API.

What it does

This workflow outlines the following key steps:

1. Listens for Form Submissions: The workflow is triggered when a user submits data through an n8n form. This form likely collects the user's query or input for the chatbot.
2. Processes User Input: The submitted form data is then passed to subsequent nodes for processing.
3. Makes HTTP Requests: An HTTP Request node is included, indicating interaction with external APIs. This would typically involve:
   • Sending the user's query to an AI model (e.g., OpenAI, custom LLM) for initial processing or response generation.
   • Querying an Infranodus knowledge graph to retrieve relevant context, facts, or relationships based on the user's input or the AI's initial response.
   • Potentially sending the combined information (user query, AI response, Infranodus context) back to the AI for a refined, knowledge-graph-enhanced response.
4. Provides a Form Interface: A standalone n8n Form node is present, suggesting a user-friendly interface for submitting queries to the chatbot.
5. Includes a Sticky Note for Documentation: A sticky note is included, which is often used for in-workflow documentation, instructions, or important notes for users or maintainers of the workflow.

Prerequisites/Requirements

To effectively use and extend this workflow, you will likely need:

• n8n Instance: A running n8n instance to host and execute the workflow.
• Infranodus Account/Instance: Access to an Infranodus knowledge graph, including its API endpoints and any necessary API keys for querying.
• AI Service API Key: An API key for an AI service (e.g., OpenAI, Google AI, custom LLM) that will process natural language queries and generate responses.
• Basic API Knowledge: Familiarity with making HTTP requests to configure the HTTP Request node for your specific AI and Infranodus APIs.

Setup/Usage

1. Import the Workflow: Import the provided JSON into your n8n instance.
2. Configure the On form submission Trigger:
   • Access the On form submission node.
   • Ensure the form is set up to collect the necessary user input (e.g., a text field for the query).
   • Activate the workflow to get the webhook URL for the form.
3. Configure the HTTP Request Node:
   • Edit the HTTP Request node.
   • AI API Integration:
     • Set the Method (e.g., POST) and URL for your chosen AI service's chat or completion endpoint.
     • Add Headers for authentication (e.g., Authorization: Bearer YOUR_AI_API_KEY).
     • Construct the Body of the request to send the user's query, potentially including system prompts or context.
   • Infranodus API Integration:
     • You might need additional HTTP Request nodes or modify the existing one to interact with the Infranodus API.
     • Determine the Infranodus API endpoints for querying your knowledge graph.
     • Add necessary authentication (e.g., API keys, tokens).
     • Construct the request body to send relevant terms or concepts extracted from the user's query or AI response to Infranodus.
4. Process and Combine Responses: Add further nodes (e.g., Code, Set) to:
   • Extract the AI's response.
   • Extract relevant information from the Infranodus API response.
   • Combine these pieces of information into a single, enhanced response.
   • Potentially send the combined information back to the AI for a final, refined output.
5. Display the Chatbot Response: Add a node to display the final, enhanced chatbot response to the user, perhaps by updating the form or sending it to another messaging service.
6. Activate the Workflow: Once configured, activate the workflow to make it live.
7. Test: Submit a query via the n8n form to test the end-to-end functionality.