Here’s a fact about my current role: creating emails is a hard process.

We use a custom built marketing platform for email campaigns, and emails are built using Handlebars.js, Node, and version controlled in a GitHub repository.

In order to create a new email, we must create a Git branch, create Handlebars.js and JSON files, and then create a Pull Request in GitHub to merge our new email into production.

Before AI was unleashed, I did everything possible to automate this process… which involved creating a lot of bash scripts to run from Terminal to streamline my production workflow.

Then AI was unleashed.

At first, we just had an in-house GenAI platform, and I started experimenting with that. It was kind of good.

Then we were given access to Google Gemini, and I started using that to refine my bash scripts to make them even more powerful and robust. I learned that AI could accurately and reliably write code, given proper instructions.

After that, AI coding assistants were made available, notably all developers could start using Cursor. And that changed everything.

With Cursor, my team has been building Cursor Rules and Commands to fully automate building marketing emails. We simply give it a few basic instructions, our marketing email brief, and then let the agent do to the heavy lifting. Because we’ve created Rules, Cursor has guidance on how to code an email template. We’ve also been leveraging Figma’s MCP connection to Cursor to streamline building new modules and components for our template system.

I never believed I’d be able to streamline email production this much without any engineering support, but that time is now here.