I decided to dive into n8n because, let’s face it, it was a massive trend this year. Everyone with a smartphone and a social media account was raving about the sheer power of this platform, claiming you could automate your entire work and daily life through it.

I had recently wrapped up a full course on Python automation, alongside an intro to AI. So, you can imagine my surprise when I discovered a platform that merges all of this into a single open source and highly customizable environment. It didn’t take long before I carved out a spot for n8n on my learning schedule.

The biggest surprise came when I started looking for learning resources. I stumbled upon their official channel, which featured two full courses packed with practical applications one for the basics and another for slightly more advanced topics. By the time I finished the first course and reached the middle of the second, I realized the whole thing didn’t quite warrant the intimidating aura around it. It wasn’t nearly as difficult as I had imagined. At its core, it was just a simple Node.js based system offering pre-built code blocks in the form of “nodes”. The entire magic simply revolved around connecting these nodes and tailoring them to fit the exact automation flow you wanted to achieve.

I quickly put together a simple workflow to add a new row to my daily tracking table on Notion. I paired that with a weekly system that generated a progress report, complete with completed tasks and percentage metrics.

Unsurprisingly, I soon hit a minor but fundamentally critical roadblock. The system wouldn’t run once I shut down my computer because it was hosted locally. The automation was practically useless if it wasn’t actually automated. I was left with two choices : either subscribe to one of n8n’s paid tiers or host it myself on an external server.

Since I’m not a fan of monthly subscriptions especially before fully utilizing a tool I leaned towards the cloud server route. I started digging, but ran into multiple walls. Free servers were scarce, the ones that existed required endless approvals, and most weren’t even supported over here.

That’s when I stumbled upon a bizarre workaround suggested by my personal AI assistant. Why not deploy it on a Docker container ? They offered a free cloud tier with around 4GB of RAM (I don’t recall the exact specs), originally intended for research, testing, and developing AI models.

Let me tell you, it was anything but easy. From the initial n8n installation to tweaking the interface just to fly under the radar and avoid getting banned, it was a grind. I later discovered that, for various reasons I couldn’t even use the latest versions of n8n, forcing me to reinstall older builds. And don’t even get me started on the SSL and data encryption headaches.

But against all odds, it finally worked. Honestly, there are no words to describe that rush of adrenaline it felt like I had just hacked into a highly secured military mainframe. The real joy though, was waking up in the following days and watching it run entirely on its own.

Of course, it didn’t take long for the platform to catch on to my little stunt. The system was shut down less than a month after I activated it, but it was an absolutely irreplaceable experience.

I eventually ended up buying a dedicated cloud server for my work and automation needs, but… that’s a another story.