Every week it was the same cycle staring at the fridge trying to figure out what to cook, heading to the market with no real plan, picking up random things, forgetting half of what I needed, and somehow still ending up with nothing that made a complete meal. It was exhausting just in that quiet background-drain kind of way that you stop noticing until someone asks you why you've eaten the same thing three times this week.

One day I asked Claude to just plan two weeks of meals for me and list everything I'd need to buy. It worked, I had a plan, a list, prices roughly estimated all in a single conversation. And then I thought: why is there no actual product that does this?

The Problem

The daily question of "what are we eating?" sounds small until you're the one answering it every single day. It drains mental energy, leads to repetitive meals, and almost always ends in an unplanned market run where you spend more than you intended because you had no list.

The tools that exist weren't built for this context. They don't know these meals, they don't price in Naira, and they don't understand a household feeding four people on a real budget.

There was a clear gap. And it was personal.

The Solution

Here's where this project is different from most case studies: I didn't hand off designs to a developer. I built it myself, using Claude Code, Anthropic's AI coding tool that works directly in your terminal and can read, write, and edit files on your machine.

I went into this as a product designer, I don't write code. But I know exactly what good looks like, I can describe what I want with precision, and I understand how products should feel. It turns out that's enough.

The process worked like this: I described what I wanted in plain language, Claude Code built it. I reviewed it as a designer what was off, what felt wrong, what was missing entirely, I gave feedback, it made changes, and we kept going, that loop is still ongoing. The product in its current state is a starting point, not a finished thing there's a significant amount of design evaluation, restructuring, and refinement still ahead. What AI accelerated was getting something real to react to. The actual design thinking, what to change, why, and how is still the work.

Chop lets users set their dietary preferences once, what they eat, what they avoid, allergies, how many people they're feeding. They pick a plan duration: one week, two weeks, or a full month. The app generates a complete daily meal plan with breakfast, lunch, and dinner built around their preferences. Then it produces a full shopping list with every ingredient, grouped by category, with estimated prices and a total budget figure. Users can swap any meal they don't feel like eating, use the app as a guest, or sign in to save and revisit past plans.

Future Plans

Chop is very much still in progress. The core flow exists, the landing page is live, but both need significant design work the landing page is a starting point, not a finished product, and the app experience is even earlier in its evolution. The next phase involves a full design pass across every screen, and letting actual feedback determine what gets prioritised next.

The goal was never to ship something perfect. It was to ship something real, learn from it, and build the next version with more information than I had when I started. That process is ongoing and that's the point.