n8n Workflow: Detect Anomalous B2B Orders in Google Sheets

Automatically flag anomalous B2B orders in Google Sheets using AI risk scoring

This n8n Workflow reviews pending distributor orders stored in Google Sheets, enriches each order with customer and order-history context, uses Google Gemini to assess anomaly risk, then updates the order status and notifies Slack for high-risk cases.

What this workflow does

• Runs every 15 minutes on a schedule.
• Pulls the first order marked as “Pending” from the Orders tab in your Google Sheets document.
• Looks up related data across sheets—customer profile, order history, product details, credit position, and recent returns—then calculates historical averages, maximums, and relevant ratios/metrics.
• Builds a consolidated context object (order + customer + history ratios + credit utilization + returns metrics) and sends it to Google Gemini for a structured risk assessment.
• Updates the order row in Google Sheets with the AI risk level, risk score, risk reason, and sets status to Approved or Under Review.
• If review is required, appends a case to the Review_Cases tab and sends a formatted alert to your configured Slack channel.
• Writes an audit log entry to the Audit_Log tab including the model name and the final decision details.

Use cases

• Detect unusual purchasing behavior in distributor orders before fulfillment.
• Catch credit-utilization spikes or return-heavy patterns using order-history metrics.
• Create an auditable review queue for finance/ops teams via Google Sheets + Slack alerts.

Technical details

• Integrations: Google Sheets, Slack, Google Gemini
• n8n nodes used: if, set, slack, summarize, sticky note, google sheets

Video demo: https://youtu.be/f7Kh1q_xL9I