← all writing

Why I built Glaner: an app that turns your fridge into recipe ideas

Glaner scans a photo of your fridge and turns what's already there into recipes, so ingredients stop going to waste. Here's why I built it and how it works.

There’s a moment most people who cook know well. You open the fridge, look at what’s inside, and none of it tells you anything useful. Half a carton of cream, three eggs, some spinach a day or two from wilting, a jar of something bought for a recipe you never got around to finishing. You close the fridge and order food instead.

That’s the exact gap Glaner is built for.

Open the app, take a photo of your fridge or pantry, and Glaner reads the ingredients in that photo and turns them into recipes built around what’s already there. No typing out a shopping list, no scrolling past ten recipes that each need three things you don’t have. You point the camera at what you own, and the app works backward from that, not forward from a recipe you found somewhere else.

Under the hood it’s a native iOS app, built in SwiftUI, using AI to identify what’s in the photo and generate recipes around exactly those ingredients, with substitutions suggested for anything missing. You can edit the detected list before generating anything, in case the AI mistakes the yogurt for something else, or misses the jar hiding behind the milk carton.

I built Glaner for a fairly ordinary reason: I kept throwing food away. Not because I dislike cooking, I don’t, but because figuring out what to make from a half-random set of ingredients takes more mental effort than most people have left at 7pm on a weekday. So things sat in the fridge until they stopped needing a decision and just needed binning.

It turns out that’s not a personal quirk. Cooking forums are full of the same complaint, worded differently each time: people tracking ingredients manually in spreadsheets or notes apps because nothing does it for them, ending up with odd, unused amounts of things going bad at the back of the fridge, trying paid meal-planning services and finding the recipes underwhelming for what they cost. The problem was never a shortage of recipes, the internet has an infinite supply of those. The problem is going from “this is what I have” to “this is what I can make” without doing that matching by hand every single time.

Glaner is built for that specific moment, not for meal planning in general and not for people who already know exactly what they’re cooking tonight. It’s for people staring into an open fridge with no plan, people trying to use things up before they spoil, and people who’d rather spend two minutes photographing a shelf than fifteen minutes scrolling recipe sites for something that fits what they already bought.

I’ve been building it solo from Setúbal, alongside a full-time job writing biometric identity SDKs, which is a completely different kind of engineering problem. The app is live now on iOS. The next thing I’m working on is the natural extension of the idea: instead of one photo producing one set of recipes, suggesting a few recipes for the week that deliberately share ingredients with each other, so that odd carton of cream from the start of this post actually gets used up, instead of forgotten again.