Well… I’m back!

It’s been a few years since I last blogged, and somehow the world went from “machine learning is neat” to “AI agents are cutting deals with each other using LLMs over WebRTC”. Although we all know AI is moving fast, but I feel we totally underestimated it. I remember tinkering with MachineBox for facial and object recognition, running Tensorflow and Llama.cpp on my Raspberry Pis before the ChatGPT came along. I figured it’s time to fire up this dusty old blog again. This time, with sharp focus on Artificial Intelligence and Machine Learning.

Over the years, I’ve spent way too many nights architecting enterprise systems, building large-scale solutions, battling cloud, legacy modernization, fighting Cyber Security fires, and giving my rack servers just enough intelligence to judge my cable management. Back in the day, I blogged about home automation and DIY projects. If you knew me from the automation world, you probably saw me haunting Discord home automation servers 24/7, programming IoT devices, writing Jinja scripts, and figuring out how to make every new device in my house talk to each other without catching fire. P.S: I’ve had my fair share of circuits blowing up on my face. What an amazing way to learn, right?

Fast forward to today: I realized most of the old blog content was… well, outdated. So, I nuked it. all my blog posts about home automation are… gone! Clean slate. Fresh start. And this time, it’s all about AI now!

Now I’m diving headfirst into the deep end - LLMs, private stacks, RAG pipelines, agents, GraphRAG, edge deployments…etc, you name it. I’ll be sharing everything I learn, build, break, fix, and occasionally rage about.

Expect thoughts, hacks, patterns, lessons learned, a few snarky takes on trending frameworks, and way too many experiments that probably shouldn’t be run in production (but might anyway).

Let the chaos (and learning) begin, and let’s fire up those GPUs!