The Newsletter Automation Trap Nobody Talks About

The #1 Mistake Most People make with
Automated Newsletters

I’ve built my career on a simple principle: if a task is repetitive, automate it. (Click here to get the full workflow.)

So you might be surprised to learn that one of the most critical parts of my content system — the part that fuels my newsletters and social media — is still manual. By choice. It’s a decision that goes against my grain of pure automation, but it’s the secret to why my system actually works.

With countless news sources to track, my first instinct was to let an AI agent handle the selection.

I built a system to automate the curation, assuming it would intelligently pick the best articles for me.

But I quickly found that in automating the curation, the most important articles weren’t getting through. The AI was efficient, but it lacked judgment.

It would consistently skip the nuanced, insightful pieces that I knew were critical for my audience. The system was filtering out the signal along with the noise.

So I made a deliberate decision: instead of total automation, I needed a human in the loop. I wanted to contribute my own creativity and have the final say on what news I believe should be highlighted.

This isn’t a failure of automation; it’s a recognition of where human expertise is irreplaceable.

Hence, I created my Augmented Curation Engine (FREE on GitHub). I used a tool called Lovable to build a custom dashboard that helps me personally select the most relevant content. AI does the grunt work of gathering, but I provide the critical layer of human insight.

Here’s how it works:

Step 1: The Automated Scout

This part is fully automated. My system constantly scans my handpicked list of top-tier tech and AI news sites. It extracts the key data from each article — title, link, content, image URL, and publication date — and standardizes it. 

This overcomes the technical challenges of encrypted content and messy formatting, creating a clean, structured feed in our database.

Get this workflow by commenting AUTOMATION

This AI scout does the grunt work of reading everything, so I don’t have to.

Step 2: The Human-in-the-Loop Control Panel

This is the manual step, and it’s the most important.

Instead of an AI agent guessing which articles are most relevant, all the scraped news flows into a simple dashboard I built. Here, I can see everything at a glance, sorted by source. With a toggle, I can condense the view to just headlines.

In under five minutes, I can scan the day’s intelligence and personally select the 5–6 articles that align with my unique perspective and what I know my audience needs. I check the boxes, press a single button, and my curated list is saved to a new table.

It ensures that every piece of content that goes out has my seal of approval. It slashed my curation time from hours to just minutes.

I used a mix of Lovable AI and Cursor to create this simple yet intuitive dashboard for me to curate the daily news items.

Step 3: The AI Co-Writer

Once I’ve made my selections, a new automated workflow kicks in. It takes my curated list and feeds each article to a GPT OSS 120B model (OpenAI’s Open Source Model — FINALLY!). This AI doesn’t just summarize; it rewrites the core ideas into distinct, on-brand content blocks, ready for publishing.

GPT OSS 120B n8n Agent

Step 4: The Multi-Channel Distributor

The final stage is publishing. Right now, the system is configured to post these curated insights as articles on Medium, a platform with a reliable API. A little note: Medium has since sacked their API sadly, however those that applied for access a while back to the API still have it. 

Anyways, the next evolution is to feed these content blocks directly into our other systems. Imagine these insights fueling Animatic to automatically generate a week’s worth of short-form videos, or being scheduled as posts across LinkedIn and X.

This is how you move from simple process automation (Level 2) to scaling an entire department’s output (Level 3). You build a system where human strategy directs AI execution.

It was tricky at first to set up the Supabase table. So if you are not tech savvy, I will have the full detailed instructions in our Corporate Automation Library (CAL) which will host the n8n code and steps required to get this running on your server. 

Click Here to get access to CAL. We plan on having over 100+ automations with 2–4 corporate automations uploaded weekly.

Want the framework to apply these ideas?

I share systems like this every week in my free newsletter. When you sign up, I’ll immediately send you my AI Automation Audit. 

It’s a simple tool to find out which of the 5 Levels of AI Automation your business is currently on, so you know exactly where to focus next.

Find Your Level & Join Here