When CampusIQ started going hard on AI, I assumed it was for the engineers. You know, the people who already lived in terminals and actually understood what all those words meant. I had never opened a terminal without someone standing over my shoulder telling me exactly what to type.

Aaron pushed me anyway, and I’ve told him off more than once for it. Always when I’m alone. In my car. In my office. During my morning shower. I’m not a developer. This is not my job. Stop making me do things I was never supposed to do.

Turns out, I was looking at it the wrong way.

I started mostly out of stubbornness, with maybe a little spite mixed in. There were parts of my job that ate entire afternoons, like pulling the same status updates from a dozen places, rebuilding the same reports every week, and stitching together client narratives by hand.

I knew those processes better than anyone because I was the one doing them over and over again. I knew where the information lived, what mattered, what could be ignored, and what the finished result actually needed to look like. I just didn’t know how to build something that could do the tedious parts for me.

AI changed that.

I still can’t write a for-loop. I still don’t understand half the words in my own pull requests, and I still picture a snake when someone says Python. But I can look at a messy, repetitive process and say, “This is dumb. Here’s what I actually need.”

It turns out that’s a pretty useful skill.

These aren’t projects an engineer would have been excited to put on a roadmap. They were small, specific problems buried inside client success. More specifically, they were my problems.

That’s what I think gets missed when we talk about non-technical employees using AI. The people closest to the work often know exactly what needs to be fixed. Before, we either kept doing it manually or had to explain the problem to someone technical and wait for them to build a solution.

Now I can build much more of it myself.

I spent years learning how Client Success works. I know what a customer risk looks like before it becomes a bigger problem. I know which information matters in a report and which information is noise. I know when a renewal conversation going quiet means something and when it doesn’t.

AI didn’t give me any of that.

What it gave me was a way to turn what I already knew into something useful without needing to become a developer first.

That distinction matters. The value I bring to these projects isn’t that I suddenly know how to code. It’s that I understand the problem well enough to describe what needs to happen, recognize when the result is wrong, and know what a useful solution should actually do.

The technical barrier got lower. The value of knowing the work didn’t.

The biggest surprise wasn’t that I could ship something into production. It was what happened to the rest of my job once I started getting repetitive work out of the way.

The status pulls, reports, and follow-ups weren’t individually difficult, but together they took up time and attention that I would rather spend somewhere else. Automating more of that work gave me more room for the parts of Client Success that still need me: judgment, relationships, and the moments when a customer needs someone who’s seen this movie before.

That’s ultimately what made AI useful to me. It wasn’t about turning myself into a technical person at 71. It was about spending less of my time on work that didn’t need 40 years of experience to get done.

I’m now shipping more code than some of our kids and grandkids, which is still objectively funny to me.

But I don’t think the lesson here is that a 71-year-old woman learned how to code. I didn’t, really.

I learned that being non-technical didn’t mean I had to accept every tedious process as something I would do by hand forever. The things I already knew about my customers, my team, and my job gave me the hardest part: understanding the problem and knowing what good looks like.

The tools don’t care how old you are.

They don’t care what your title is.

They only care whether you’re willing to describe the problem clearly and stop doing the tedious work by hand.

So yes, the old dog learned some new tricks. But she didn’t have to become a different dog to do it.