She Gave Her AI a Job Title. Now It Runs Her Household.
So I came across this post from Lauren Selley that hit way too close to home.
Lauren is a systems designer by trade. She builds processes and repeatable workflows for other people's businesses all day. Then she goes home to a three year old who only eats beige food and a newborn, and suddenly the simplest decisions turn into hour long research spirals. Comparing kids' shoes for 45 minutes. Reading rug reviews until her eyes glaze over. Twenty minutes squinting at ingredient labels on a $4.99 snack.
Sound familiar? Yeah, me too.
Her problem wasn't that she was bad at decisions. It was the opposite. She's so good at building systems that she couldn't stop herself from treating every household purchase like a full research project. Every product category meant rebuilding context from scratch, reading affiliate blogs written by people who clearly never touched the product, and drowning in conflicting reviews.
So she did something that I thought was brilliant. She built a custom AI prompt and called it her "Household COO."
Not a casual "hey ChatGPT what's for dinner" kind of thing. An actual operational brief. She loaded it up with standing context about her family (who they are, where they live, brands they like, what matters to them in a purchase). She built in decision logic so small purchases get a direct recommendation while big ones get a structured intake process. She set up a source hierarchy that prioritizes long term owner reviews and enthusiast communities over sponsored content. And she defined an output format: comparison tables, pros and cons, a declared winner with reasoning.
The result? A rug purchase that used to take 45 minutes of Amazon scrolling and Reddit threads now takes under ten minutes. With better information than she had before.
What got me about this story is that Lauren didn't learn to code. She didn't build an app. She just thought about her problem the way she thinks about her clients' problems and applied the same framework to her own life. She treated the AI like a new contractor: give them the background upfront, define the deliverable, set the quality bar.
This is the stuff I think about constantly. When I built PillStreak for my sister in law who's in nursing school, it started the exact same way. She had a problem (memorizing drug names). I had access to tools that could help. The gap between "I have a problem" and "there's a solution" keeps getting smaller.
And that's really the point. You don't need to be technical to build something useful with AI. You just need to be specific about what you need and willing to experiment a little. Lauren's a systems thinker who happens to be a mom. I'm a former nightclub ops guy who happens to build web apps now. The tools don't care about your resume.