How can I use an n8n AI Agent to split input data into separate Google Sheets columns?

Below is a step-by-step approach to resolve the issue:

1. Instruct the AI to Output Structured Data

Instead of having GPT-4 return plain text, format your prompt so that the response is in a structured (JSON) format. For example, you could ask it to return an object like:

JSONCopied
{
  "date": "YYYY-MM-DD",
  "description": "Expense description",
  "amount": "123.45"
}

This ensures that when the AI responds, each piece of information is clearly separated.

2. Parse the AI Response in n8n

Once your AI Agent node outputs the structured response, you can use a Code node (or a Function node) to parse it. For example:

JAVASCRIPTCopied
// Assume the AI's output is in the field "text" as a JSON string
const data = JSON.parse(items[0].json.text);
return [{ json: data }];

This converts the JSON string into an object where each property (date, description, amount) is a separate key.

3. Map the Parsed Data to Google Sheets

When using the Google Sheets node to insert a row, use the parsed JSON keys as the columns. In the Google Sheets node’s parameters, you can map:

• Date → {{$json.date}}
• Description → {{$json.description}}
• Amount → {{$json.amount}}

This way each piece of your expense data will be inserted into its respective cell instead of a single cell.

4. Tutorials and Guides

To help you get started with n8n AI workflows and improve your automation skills, the following resources might be useful:

• Tutorial: Build an AI workflow in n8n
  This step-by-step guide walks you through creating workflows involving the Chat Trigger and AI Agent nodes. It covers concepts like setting up an agent, connecting to models, and processing responses.

• Chat with a Google Sheet Using AI
  This example shows how to interact with Google Sheets using an AI Agent. Although the focus is on a chat interface, it demonstrates the integration between AI outputs and structured data processing.

• n8n Docs on AI Agent/Cluster Nodes
  Check out the documentation for detailed insights into using the AI Agent node and associated sub-nodes (for example, Custom n8n Workflow Tool) that help you connect different parts of your workflow.

Summary

1. Format your AI prompt to ensure a structured (JSON) response.
2. Parse the AI response using a Code/Function node to separate your expense data.
3. Map each parsed field to its respective cell in the Google Sheets node.
4. Explore n8n’s tutorials and documentation to strengthen your overall workflow automation skills.

By following these steps, you should be able to have your expense data correctly split into the appropriate columns in Google Sheets. If you need more detailed guidance on any of these steps, please refer to the specific tutorials or let me know which part you’d like to dive deeper into.