Automate Bank Statement and Invoice Reconciliation with GPT-4.1-mini and Google Sheets

🟢 Manual Trigger

Workflow starts manually to initiate the reconciliation process on demand.

📄 Fetch Invoices & Bank Statements

Retrieves invoice data and bank statement data from Google Sheets for comparison.

🔀 Merge Data

Combines both datasets into a single structured dataset for processing.

🧩 Format Payload for AI

Function node prepares and structures the merged data into a clean JSON payload for AI analysis.

🤖 AI Reconciliation

AI Agent analyzes the invoice and bank statement data to identify matches, discrepancies, and reconciled entries.

🧮 Parse AI Output

Parses the AI response into a structured format suitable for adding back to Google Sheets.

📊 Update Sheets

Adds the reconciled data and reconciliation results into the target Google Sheet for recordkeeping.

🧾 Prerequisites

✅ OpenAI API Credentials Required for the AI Reconciliation node to process and match transactions.

Add your OpenAI API key in n8n → Credentials → OpenAI.

✅ Google Sheets Credentials Needed to read invoice and bank statement data and to write reconciled results. Add credentials in n8n → Credentials → Google Sheets.

✅ Google Sheets Setup The connected spreadsheet must contain the following tabs: Invoices – for invoice data Bank_Statement – for bank transaction data Reconciled_Data – for storing the AI-processed reconciliation output

✅ Tab Structure & Required Headers

Invoices Sheet Columns: Invoice_ID Invoice_Date Customer_Name Amount Status

Bank_Statement Sheet Columns: Transaction_ID Transaction_Date Description Debit/Credit Amount

Reconciled_Data Sheet Columns: Invoice_ID Transaction_ID Matched_Status Remarks Confidence_Score

⚙️ n8n Environment Setup

Ensure all nodes are connected correctly and the workflow has permission to access the required sheets.

Test each fetch and write operation before running the full workflow.