Student Motivation RAG Pipeline with Weaviate & OpenAI

Overview

This workflow implements a Retrieval-Augmented Generation (RAG) pipeline that processes incoming webhook payloads (e.g., daily motivational prompts) by splitting text, embedding it with OpenAI, and storing/querying vectors in Weaviate. The RAG Agent then generates a response using an OpenAI chat model, logs the result to Google Sheets, and sends a Slack alert on errors. It's designed to automate a "Daily Student Motivation" system where an external scheduler or user sends a request via webhook.

Node-by-Node Breakdown

• Webhook Trigger (No auth) — Listens for POST requests at /daily-student-motivation. The webhook URL must be configured in n8n; external callers send JSON payloads containing the motivation content.
• Text Splitter (No auth) — Splits incoming text into chunks of 400 characters with 40-character overlap. This improves embedding accuracy by breaking long texts into manageable pieces.
• Embeddings (API Key auth — OpenAI) — Uses the text-embedding-3-small model to convert text chunks into vector embeddings. Requires an OpenAI API key credential.
• Weaviate Insert (API Key auth — Weaviate) — Inserts the generated embeddings into a Weaviate index named daily_student_motivation. Requires Weaviate credentials (host URL, API key).
• Weaviate Query (API Key auth — Weaviate) — Queries the same Weaviate index to retrieve relevant vectors. It uses the same embedding model (connected via the Embeddings node).
• Vector Tool (No auth) — Wraps the Weaviate vector store as a tool for the RAG Agent. This allows the agent to call vector search during response generation.
• Window Memory (No auth) — Provides conversational memory (buffer window) so the agent can maintain context across multiple webhook calls (if invoked repeatedly).
• Chat Model (API Key auth — OpenAI) — Uses an OpenAI chat model (e.g., gpt-4 or default) to generate responses. Requires the same OpenAI credentials as the Embeddings node.
• RAG Agent (No auth) — The core agent that orchestrates the pipeline. It takes a system message "You are an assistant for Daily Student Motivation" and a user prompt that includes the raw JSON input. It uses the Vector Tool for retrieval, Window Memory for context, and the Chat Model for generation.
• Append Sheet (OAuth2 — Google Sheets) — Appends the agent's output (field text) to a specified Google Sheet (sheet ID SHEET_ID, sheet name Log). Requires Google Sheets OAuth2 credentials.
• Slack Alert (OAuth2 — Slack) — On error (connected via onError), sends a message to the #alerts channel with the error details. Requires Slack OAuth2 credentials.

Setup Instructions

1. Create credentials in n8n for:
   • OpenAI API: Obtain API key from OpenAI platform.
   • Weaviate: Set up a Weaviate instance (cloud or local) and get the host URL and API key.
   • Google Sheets: Enable the Sheets API in Google Cloud Console and obtain OAuth2 credentials (client ID/secret) or use a service account.
   • Slack: Create a Slack app with chat:write scope, install it to your workspace, and get an OAuth token.
2. Configure the webhook URL – after importing the workflow, note the generated webhook URL (e.g., https://your-n8n-instance/webhook/daily-student-motivation). You can call it from an external scheduler (e.g., cron job) or manually send a POST request with a JSON body containing the daily motivation text.
3. Set the Google Sheet ID – replace SHEET_ID in the Append Sheet node with your actual Google Sheet ID. Make sure the sheet has a tab named Log with a column Status.
4. Test – send a sample POST request to the webhook with payload like {"text": "Today's motivation: Stay curious and keep learning!"}.

Use Cases & Variations

• Daily Motivation for Students: Use this workflow to automatically process and store motivational messages, making them searchable via the vector store and generating contextual responses.
• RAG over Custom Content: Replace the webhook with a scheduled trigger (e.g., n8n Cron node) to ingest content from RSS feeds, emails, or databases into Weaviate for later retrieval.
• Alternative Outputs: Instead of Google Sheets, log to Airtable, Notion, or send an email via SMTP. You can also add a Slack success message by connecting a second Slack node to the main output.
• Error Handling: The current error path sends to Slack; you could also log errors to a file or a database.
• Memory & Context: The window memory can be adjusted or replaced with a longer-term memory (e.g., Redis) for more persistent conversations.

{
  "name": "Daily Student Motivation",
  "nodes": [
    {
      "parameters": {
        "content": "Automated workflow: Daily Student Motivation",
        "height": 530,
        "width": 1100,
        "color": 5
      },
      "id": "26995548-761c-41fa-9727-9c1538208723",
      "name": "Sticky Note",
      "type": "n8n-nodes-base.stickyNote",
      "typeVersion": 1,
      "position": [
        -480,
        -240
      ]
    },
    {
      "parameters": {
        "httpMethod": "POST",
        "path": "daily-student-motivation"
      },
      "id": "b4ee6923-9695-4d25-879e-d51682b416d6",
      "name": "Webhook Trigger",
      "type": "n8n-nodes-base.webhook",
      "typeVersion": 1,
      "position": [
        -300,
        0
      ]
    },
    {
      "parameters": {
        "chunkSize": 400,
        "chunkOverlap": 40
      },
      "id": "950769fc-8534-4ffa-9fad-8c9cee196dea",
      "name": "Text Splitter",
      "type": "@n8n/n8n-nodes-langchain.textSplitterCharacterTextSplitter",
      "typeVersion": 1,
      "position": [
        -130,
        0
      ]
    },
    {
      "parameters": {
        "model": "text-embedding-3-small",
// ... truncated (copy to see full JSON)