A behind-the-scenes look at my ongoing experiments in AI and automation.This is where ideas get tested, agents get trained (or confused), and workflows slowly start running themselves - kind of. I'm building tools that help with creative, research, and everyday tasks, all while figuring out what's actually useful and what's just… entertainingly weird. Project X is less of a polished showcase and more of a working lab - the kind with digital clutter, half-finished scripts, and lots of coffee. It's where I hand over the boring bits to AI and see what happens next. Follow along if you're curious about the messy, real side of reskilling, experimenting, and learning by doing.
Early version of the automation dashboard prototype.
What started as a mini side project where I was trying to make my own content and marketing production workflows more efficient has turned into a bit of an unruly lab for all things AI and automation.
The first AI agent (two years in the making) is shaping up to support my creative, research and content work, with a few coding tasks thrown in. Around it, a handful of bots and workflows are quietly helping with everything from digital admin (like tidying up and organising my desktop) to the occasional oddball experiment (I need one that can create recipes out of leftover food). Time is life's most precious commodity so if I can automate something, I probably will!
I'm exploring ways to streamline content creation and digital workflows using tools like Python, LangChain, HuggingFace, agent frameworks and different LLM APIs - testing out agents, prompt engineering, and local-first automations.
This project grows as I learn, with plenty of trial, error, and “wait, that actually worked?” moments along the way.
As I build, break and learn things, I'll be posting updates so you can follow along. That way it keeps me honest. And feel free to reach out if this sparks your interest-I'm always up for a good chat about all things AI without the tech jargon.