The Machines Were Never the Point
Long before AI, I was the kid taking the family computer apart. The obsession never left — it just found a bigger machine to point at.
I was ten the first time I opened a computer case on purpose — not to fix anything, just to see what was actually in there. I was learning C++ at NIIT around the same time, understanding maybe half of what I typed and loving all of it anyway. Two years later I sat the Microsoft Certified Systems Engineer exam, built for adults running corporate networks, and passed it. At home this made me the kid everyone called when Windows 95 did something inexplicable — other people saw a computer breaking, I saw a system I hadn’t fully mapped yet.
That instinct didn’t become a career, at least not directly. The next two decades went into audit, restaurants, cloud kitchens, capital raises — good work, real work, but not the work of taking a system apart to see what was inside. The instinct wasn’t gone. It just didn’t have a machine worth opening.
Then AI happened, and it came back immediately. I went through models and tools the way I used to go through a PC case — compulsively, trying to find the edge of what each one could actually do versus what it claimed to do. Somewhere in the middle of testing everything against everything, the question stopped being what can this do and became what can I build with it. That compulsion is the reason ClarityHQ exists.
Four of those builds are live today, run by real businesses, each written up properly: a commerce backbone for Travaholic, live in two days instead of two months; a direct-booking system for Travaholic Stays, replacing Airbnb and a spreadsheet; a full digital operation for The Feeling Co. running on zero external subscriptions; and a two-sided marketplace connecting Indian artists to European collectors. Four different problems, one instinct behind all of them: open the thing up, understand every part, put it back together better.
Twenty years and one technological revolution didn’t change the actual thing: the machine was never the point. At ten, the interesting part wasn’t the motherboard — it was understanding it well enough that it stopped being a black box. Now the interesting part still isn’t the model. It’s knowing exactly where to point it, and where not to. What was scarce when I was ten — someone who actually understood the system instead of just using it — is still scarce now. I just get a bigger machine to work on.
If you’re building something and want a second, technically fluent set of eyes on it, get in touch.