Kieron Harper

I started out doing what most developers do — building websites and internal tools, fixing bugs, shipping features. Good work, but nothing that felt like it had real leverage. I was always looking for the thing that multiplied output rather than just adding to it.

The shift happened gradually. I started experimenting with LLM APIs mostly out of curiosity, building small scripts to automate the repetitive parts of my own workflow. The results were uneven but the potential was obvious. I started spending more time there — reading papers, reverse-engineering what worked, shipping things that weren't meant to be products just to see how far the tools would stretch.

At some point the experiments became projects, and the projects started generating actual results. Time saved across half a dozen workflows. Those outcomes changed how I thought about where to put my effort.

I'm not an AI researcher. I don't work at a lab. I'm a builder who found that applied AI — the unglamorous, prompt-engineering, glue-code kind — is one of the highest-leverage things I've ever worked with. Harper Tech Lab is where I share what that actually looks like in practice.