AI Assistant for Spreadsheet Formula Generation and Explanation - n8n Workflow

Data Analysts & Finance Professionals: Who frequently need to construct or understand complex spreadsheet functions (Excel, Google Sheets).
Developers & Integrators: Looking for robust n8n templates that showcase sophisticated AI routing and web service integration.
Automation Specialists: Seeking to deploy custom n8n workflow solutions accessible via API endpoints.
Support Teams: Needing an automated tool to quickly provide documentation or solutions regarding data manipulation.

Generating precise Excel or Google Sheets formulas can be time-consuming, especially when dealing with complex data structures. This advanced n8n workflow solves this by utilizing the power of a large language model (LLM), accessed via OpenRouter (like DeepSeek, Llama, or others), to act as a specialized formula expert.

The primary value of this n8n automation is its dynamic routing capability. When the n8n trigger receives a request, the Intent Switch node intelligently determines if the user wants to generate a new formula from a description or explain an existing formula.

This architecture ensures that the AI is used efficiently, providing targeted, instant answers and demonstrating a powerful application of modern n8n templates for business logic.

This n8n workflow initiates its process using a Webhook n8n trigger, making it easily callable from any application or service.

1. Trigger & Data Intake: The Webhook n8n trigger starts the execution upon receiving an HTTP request containing the user's query (e.g., 'Generate an IF statement to check column A').


2. Intent Classification (Set Node): The Detect Intent n8n node prepares the input data, structuring the query for downstream processing and intent analysis (though intent logic might be external or configured within the next node).


3. Flow Routing: The Intent Switch n8n node is the central control point. It inspects the request payload to determine the required action. The workflow routes the request down one of two paths:

1. AI Processing (LangChain Agent Nodes): Both formula paths use dedicated LangChain Agent n8n node setups, leveraging separate instances of the OpenRouter Chat Model n8n node (configured, likely, for high-performance models like DeepSeek). These agents fulfill the specific request, either writing the formula or providing a detailed explanation.


2. Output & Cleanup: The result from the chosen AI path flows into a Code n8n node for final data formatting and cleanup.


3. Response: The final formatted output is immediately returned to the calling application using the Respond to Webhook n8n node, completing the entire n8n workflow process swiftly.

To set up this dynamic n8n template, follow these steps:

1. Import the n8n Workflow: Copy the provided JSON code and import it directly into your self-hosted or cloud n8n instance using the 'New' -> 'Import from JSON' feature.


2. Configure Credentials:

1. Configure Webhook n8n Trigger: After importing, activate the Webhook n8n trigger node. Copy the test or production URL provided by the n8n node. This URL is your endpoint for sending formula requests.


2. Define Intent Logic: The core functionality relies on the Intent Switch node. You must ensure the preceding Detect Intent n8n node or your external calling application sets a clear key/value pair (e.g., intent: 'generate' or intent: 'explain') that the Switch node can reliably check.


3. Activate: Save the n8n workflow and set the toggle to 'Active' to start listening for requests.

Webhook (n8n Trigger):
Function: Serves as the initiation point for the n8n workflow, exposing an HTTP endpoint to receive user queries.
Key Configuration: Uses a unique Webhook ID to define the accessible endpoint.
Detect Intent (Set n8n node):
Function: Used to standardize the incoming data payload and prepare the intent classification fields before routing.
Intent Switch (Switch n8n node):
Function: The core routing logic. It evaluates the input and directs the flow to either the formula generation path or the explanation path, maximizing the utility of this n8n template.
OpenRouter Chat Model / OpenRouter Chat Model1 (LLM n8n node):
Function: Provides the connection to the large language model API (via OpenRouter), enabling access to advanced models like DeepSeek for specialized text generation tasks.
Key Configuration: Requires a configured OpenRouter API credential and specific model selection (likely a high-context, reasoning model).
Generate Formula / Explain Formula (LangChain Agent n8n node):
Function: These agents orchestrate the interaction with the LLM, providing specific prompts and context (tools) tailored for either generating new Excel/Sheets formulas or describing the provided ones.
Code (n8n node):
Function: Performs final transformation, filtering, or cleanup of the AI-generated text before sending the result back.
Respond to Webhook (n8n node):
Function: Sends the AI's final answer back to the service that initiated the n8n trigger, concluding the request cycle.