Why I Built Dishcount

I'm going to tell you the moment that started all this. It's embarrassing, but it's also the reason Dishcount exists.

I was standing in a Kroger aisle about a year and a half ago, holding a $7.99 jar of something I can't even remember now. Capers? Tahini? Some condiment the recipe needed. One tablespoon of it. And I was going to buy the whole jar.

I looked at my cart. It was the usual haul, chicken, produce, pasta, yogurt for the kids. Almost none of it was on sale. I'd planned the week's meals on Sunday night, made a list, walked in, and just grabbed things off the shelves like I always did.

It hit me standing there: I'd spent twenty minutes picking recipes and exactly zero minutes checking what was on sale. And I'd been doing it that way for years.

The thing I couldn't find

That night I sat down to look for a tool that did what I wanted. I figured if I was this dumb about grocery shopping, somebody smarter had already solved it. There must be an app where you plug in your zip code, see what's actually on sale nearby, and get recipes built around the deals.

Nope.

Every meal planning app I tried worked the same way: pick your recipes first, then it makes you a shopping list. Some of them let you import coupons. A couple had vague "budget" filters. But none of them started with the sales.

And that was the whole problem. The price of your groceries isn't set by the recipe — it's set by what the store decided to put on sale that week. If you're not checking the sale first, you're literally leaving money on the shelf.

I am not a developer

Here's the part where I should say something like "so I hired a team" or "so I raised some money." I didn't. I don't have either of those things and exactly zero experience writing code or developing a website. I work a normal job. I'm in the Navy Reserve. I have a family. And I have about an hour every night, after everyone else goes to bed, to work on whatever project I thought would take 2 days but is now in week 3!

What I did have was an AI coding assistant.

I started asking Claude questions like "how would I pull grocery store weekly ads" and "what's an API" and "how do I make a website." For the first couple of weeks I didn't know what I was doing at all. I broke things constantly. I'd describe what I wanted, get code back, copy-paste it, watch it fail, and then ask why it failed. Over and over. Night after night.

Slowly, it started working. First a simple page that pulled Kroger deals by zip code. Then ALDI. Then a way to generate recipes from the sale items using AI. Then the rest.

I want to be clear: I didn't build this the "right" way. A real developer would look at the code and have opinions. But it works, and it works for real families. And that's what matters.

What it does now

You type in your zip code. Dishcount pulls the live weekly deals from Kroger, ALDI, Walmart, and about 80 other grocery chains near you. You pick the deals that look good to you. Add any ingredients you already have at home. Mention any dietary stuff like "no pork" or "dairy-free." Hit a button.

It generates meals around those deals. Full recipes. A shopping list. If you're a Kroger shopper, it can even send the list straight to your cart for pickup.

That's it. There's no app to download. No sign-up required to try it. No "premium plan." It's free because I wanted to build the thing I'd needed, and then it felt weird to charge people for it.

What I've learned

Two things, mostly.

First, the barrier to building something has never been lower. If there's a problem in your life that you wish somebody would solve, and nobody has, the tools exist right now to let you try. I'm not saying it's easy. It was hard. But it was possible, and a year ago I wouldn't have believed that.

Second, the best time to build something is when nobody's watching. No investors. No deadlines. No pressure to monetize. Just a problem you cared about and the stubbornness to keep poking at it until it worked. That's not a bad way to spend your evenings.

If Dishcount helps your family even a little, that's what it was for. And if you've got feedback — good, bad, or "hey this is broken" — I'd love to hear it. There's a contact form on the site and it goes straight to me.