From hobby baker to app builder

Saturday, 5 July, 2025

I cook a lot, bake a lot. Every time I spend a lot of time with a calculator, scaling recipes up and down. Until one Saturday morning when I realized – why am I doing this math shit when I could be building an app for it?

From hobby baker to app builder

The 15 minutes of baking fame

Picture this: my recipe for pizzas is based on Sandra Mastios “The One”. It’s 12 pieces of dough, 240g each. Easy peasy … but now I’m invited to a restaurant to make pizzas for an event where they will serve 5 different small dishes on their patio. Guests? anything between 50 and 150 people. To top it off the restaurant has a green Michelin star so regular large sloppy pizzas won’t cut it. After a bit of testing and discussiong with the chefs we decide to go for 50g size pizzas.

Now the 12 piece 240g dough recipe has to be scaled up and down to 50-150 50g doughs. wtf?

I’m standing in my kitchen knowing I need to make at least 3 test batches before the event. Each batch takes 5 days of dough fermentation, has to be identical so I first can test size and baking time, then toppings, flavour and estetics. Lastly the third test batch will be a validation that I can repeat the pizzas. So I need at least a month to be sure I can produce a solid recipt for the event.

I think you get the idea. There is no way I can rely on a calculator and manual calculations for this.

Base recepie:

Total dough weight: 2898g

Converting that to 50g doughballs will give about 57 doughs. To get the right amount of flour for 150 dough balls I need to … (1360/50)*150, and so on, through six ingredients for each test batch … bloody hell, we can all see that this won’t work and it will be a huge source of errors.

That’s when it hit me: I’m a ux designer. I solve problems for a living. Why am I manually doing arithmetic like it’s 1999?

The usecase is bigger then I thought

As I stood there, calculator in hand, I realized this wasn’t just about one recipe. I have dozens of recipes: milk bread, foccacias, pizza doughs, pixel bread, surfer buns, drunk irishmen loafs. Sometimes I want to make 5 loaves, sometimes just 10 small rolls. Every single time, it’s the same tedious math.

Describing the solution

So I opened Claude and said:

“I want an app that I can run in the browser. It shall take my bread recipes and recalculate all the ingredients when I change the batch size. Use this pizza recepie as the base.”

That’s it. No technical specifications, no user stories, no wireframes. Just a clear description of the problem I wanted to solve.

Claude asked a few clarifying questions:

Within 15 minutes, we had a woking app.

The simplest solution that could possibly work

The app Claude built is beautifully simple:

My base recipe looks like this in the .json-file:

 «name»: «Sandra Mastio - The One»,
«pieces»: 12,
«weight»: 240,
«ingredients»: {
    «Fresh Yeast»: 3,
    «water»: {
        «Cold Water»: 1080
    },
    «flour»: {
        «Tipo-00 Flour»: 1360,
        «High Protein Flour»: 340
    },
    «Salt»: 65,
    «Olive Oil»: 50
},

The app in action

[Screen rec]

Now, when I want to scale that recipe: 1. I open the app in my browser 2. Select «Cranberry Sourdough» from the dropdown 3. Change the quantity from 30 to 50 4. Hit enter

I can even add ingredients on the fly if I want to experiment with additions like seeds or nuts.

From tedious task to time saver

Building and styling this app took less than an hour. Using it saves me time and eliminates the frustration and errors of manual calculation.

But more importantly, it proved a concept: I could identify a real problem in my life and solve it with custom software. Not «find a workaround» or «live with the inconvenience,» but actually solve it completely.

That realization was intoxicating.

The Gateway Drug

This simple baking calculator became my gateway drug to building custom tools. It showed me that the barrier between «person with problem» and «person with custom solution» was much lower than I’d imagined.

If I could solve my weekend baking math in under an hour, what other annoying tasks could I eliminate from my life?

Next up: How I turned the most dreaded part of freelancing – writing proposals – into a 45-minute task instead of an 8-hour ordeal.

That question led to the proposal generator that saves me 7 hours per proposal, the image bank that eliminated my WeTransfer dependency, and the accounting tool that makes quarterly bookkeeping almost pleasant.

Almost.

The Swedish Efficiency Test

There’s something deeply satisfying about this approach that appeals to my Swedish sensibilities. Instead of accepting inefficiency or hunting for commercial solutions that almost-but-not-quite fit my needs, I built exactly what I wanted.

No subscription fees, no feature limitations, no privacy concerns about my recipes living in someone else’s cloud. Just a simple tool that does exactly what I need, nothing more, nothing less.

But here’s the kicker – it’s not just tedious, it’s error-prone. Miss a decimal point, and you’ve got 20 kilos of flour instead of 2. (Don’t ask how I know this.)

The Unexpected Benefits

What started as a simple recipe calculator turned into something much more useful:

Recipe Development: I can easily test different ratios by scaling to 1 piece and seeing the per-unit amounts.

Shopping Lists: The app shows me exactly what I need to buy, scaled to any quantity.

Consistency: No more «was it 1200g or 1260g of water?» moments. The ratios are always perfect.

Iteration: I can save variations of recipes as I experiment. My pizza dough now has three versions: thin crust, thick crust, and «experimental Saturday.»

The Learning Curve (Spoiler: There Wasn’t One)

Here’s the remarkable thing: I didn’t need to learn baking science or advanced programming to build this. I just needed to clearly articulate what I wanted.

Claude handled all the technical details: - How to parse the yaml files - How to perform the calculations
- How to handle edge cases (what if someone enters 0 pieces?) - How to make the interface responsive

My job was simply to test it and say «yes, that works» or «actually, could you make this small change?»

I’m standing in my kitchen, squinting at a recipe for cranberry sourdough that makes 30 pieces, but I need 50 for a gathering. Out comes the calculator, and I start the familiar dance of multiplication and division.