Subscribe

How I realized AI automation is all about what you automate

tl;dr: I read some of Elena Alston's articles on Zapier's blog and loved the CRM automation call example. That seamless integration of AI after a sales call really blew me away. This is what makes learning about AI automation so fun!

Yesterday, I read the first five articles of How I uncovered Zapier's best AI automation articles from 2025 with LLMs, all written by Elena Alston.

These articles are packed with a tremendous number of AI automation examples. They make you think: AI automation will be everywhere, and that's truly exciting.

In Why pairing AI with automation will change how you work ↗ I found my favorite example:

Picture this:

You attend a call with a potential prospect who shows interest in your product.

  1. After you hang up, AI sifts through the conversation, pulling out key details like:

    • pain points and

    • specific needs.

  2. The information is then automatically logged in your CRM.

  3. A personalized outreach email is drafted based on the lead's preferences and behavior.

  4. Your sales team in Slack receives a notification.

Finally the right team member can trigger the outreach.

Quite compelling AI automation, and beautiful at the same time. Right?

Looking at these real-world applications, I realized something important about AI automation engineering. The real value—and difficulty—lies not in how you automate, but in what you choose to automate: which parts of a business process can and should be automated with AI.

This got me thinking about automation. I always thought it was just about using technology for repetitive work. Was I missing something? According to Perplexity:

Automation is the use of technology to perform tasks with minimal or no human intervention, typically by making processes, systems, or apparatuses operate automatically. Automation is achieved using a range of technologies—including software, robotics, machines, and control systems—to monitor, control, and execute activities faster, more efficiently, and with fewer errors than manual processes.

In other words, Automation is removing human intervention from systems. That's broader than I used to think, for sure.

Another point I hadn't considered before: automation is especially valuable for tasks prone to human error. Think of:

  • Data entry

  • Invoice processing and payment

  • Scheduling and calendar management

  • Email sorting and responses

  • Data backup

  • etc.

I'll definitely keep this in mind.

That's all I have for today! Talk soon 👋

P.S. Thanks, Elena, for these insightful articles.

Recent posts
latestHow I crafted TL;DRs with LLMs and modernized my blog (part 5)
See how impressed I was by GPT-4.1's meta descriptions
prompt
How I crafted TL;DRs with LLMs and modernized my blog (part 4)
Check how I optimized images for better blog performance
code
How I explored Google Sheets to Gmail automation through Zapier before building it in Python (part 1)
See how I built my first Zapier Gmail alert from Sheets updates
no-code
How I uncovered Zapier's best AI automation articles from 2025 with LLMs (part 1)
See how I let LLMs rank Zapier's top 10 AI automation articles
misc
Curious about the tools I use?