How I Built a Timezone-Aware Email Automation Using n8n + OpenAI

I recently built an automated email outreach system using n8n + OpenAI, and it only sends emails during the lead's local business hours.

Here’s what the flow looks like:

This is the code used to check for the

return items.map(item => {
  // Use timezone from assistant's response (e.g. item.json.message.content)
  const timezone = item.json.message?.content || "UTC";

  try {
    const now = new Date();
    const formatter = new Intl.DateTimeFormat('en-US', {
      timeZone: timezone,
      hour: 'numeric',
      hour12: false,
    });

    const parts = formatter.formatToParts(now);
    const hour = parseInt(parts.find(p => p.type === 'hour')?.value || "0", 10);

    return {
      json: {
        ...item.json,
        currentHour: hour,
        isWithinBusinessHours: hour >= 8 && hour < 18,
        timezoneChecked: timezone
      }
    };
  } catch (err) {
    return {
      json: {
        ...item.json,
        error: `Invalid timezone: ${timezone}`,
        currentHour: null,
        isWithinBusinessHours: false,
        timezoneChecked: timezone
      }
    };
  }
});

current time `

Why I Built This

Sending cold emails at 3 AM in the recipient's timezone? Bad idea.

This automation ensures:

• Higher open rates
• Professional timing
• Less chance of being marked as spam

Tools Used

• n8n – Automation logic and orchestration
• Google Sheets – Where I store leads
• OpenAI GPT – Detect timezone + generate emails
• Custom Delay Logic – Wait until local business hours

Workflow Steps

1. Read leads from Google Sheets
2. Check status (if they’ve already been emailed)
3. Use OpenAI to infer the timezone (from country or city)
4. Check the current hour in their timezone
5. If too early/late, wait for 3 hours
6. Once it’s a good time, generate a personalized email
7. Send it, log it, and notify me of errors

Output

• Emails only go out between 9 AM–5 PM local time
• Each one is AI-personalized
• Everything is logged in Sheets
• I get notified once there is an error

Feedback?
I would love to hear:

• How do you scale or modularize this
• How to make it more robust for thousands of leads
• If there's a better approach for timezone detection

Let’s connect!
Thanks for reading