The challenge
Over the past few years, I’ve spent countless weekends perfecting my sourdough. I’d go through notebooks of ratios and hydration notes, trying to nail that balance between flour, water, starter, and salt.
The problem? Every bake started with math. How much water for 72% hydration? If I add starter, how do I account for the flour and water it already contains? If I want four pizza balls at 270 grams each, how does that scale?
Every session began with calculations, and every time, I wondered, “Why am I doing this by hand?”
The “Design” workflow looks a lot different nowadays. Gone are the days where just Figma was open.
The app icon? A picture of my own baked loaf turned into digital asset
From baking to building
As a product designer, I’ve always been drawn to the intersection of craft and tools. Baking bread is a craft. Building products is a craft.
That’s been shifting. With AI and vibe coding tools, the wall between product and engineering has never felt thinner.
Design philosophy
From the start, I challenged myself to design the Dough Calculator differently from the existing tools. Most calculators rely on fields and manual input, but as a baker, I don’t want to “fill a form;” I just want to bake.
So, I built the entire interface around sliders. You can adjust flour, hydration, or starter in a fraction of the time it normally takes to type values, which makes a big difference when you’re mid-recipe and need to update quantities quickly.
The controls are also anchored at the bottom of the screen, so your hand never covers the ingredient list. As you slide your finger, you instantly see the effect on the recipe above — no scrolling, no guesswork.
Baking has few ingredients, so I leaned into that simplicity with subtle colour accents as a nod to the physical world: white for flour, blue for water, yellow for yeast, and pink for salt (my ode to Himalayan pink salt). To keep the main experience clear, I tucked away rarely adjusted options into a secondary settings page, giving the main screen as much real estate as possible.
The result is an app that feels more like an extension of the baking process: fast, tactile, and visual, rather than a calculator pretending to be a recipe form.
Prototyping with AI
I started with ChatGPT 5, throwing it a PRD for my dream dough calculator. Within minutes, I had a working prototype: a web app with sliders for flour, hydration, and yeast type that instantly recalculated ingredients.
From there, I moved into Windsurf and Claude 4 to make the prototype more robust. This wasn’t just snippets of code; it was an end-to-end build experience: design, logic, and implementation all happened in the same flow.
Going native
Once I had a working web app, I wanted the real thing: a native iOS app. That’s where things could have gotten harder, but they didn’t.
Xcode, provisioning profiles, and bundle IDs were all new. But instead of searching through outdated tutorials, I asked GPT-5 to walk me through it live. It became an interactive coach: “Click here, check this box, now run this build.”
It reminded me of how I used to learn design tools years ago, following tutorials line by line. Except this time, the tutorial was tailored to my exact project, in real time. And the effort wasn’t wasted on a sandbox; it went straight into the product I’d ship.
Reflection
AI hasn’t replaced engineers. But it has made non-engineers feel superhuman.
Product people can bring their vision to life for the first time without waiting in line for engineering resources. We can design it, spec it, and now build it ourselves. That doesn’t diminish the role of engineers; it just expands what’s possible for everyone else.
If you’re a product person and you’re not experimenting with building right now, you’re shooting yourself in the foot. The tools are here, the workflows are real, and the gap between vision and execution is collapsing.
Download the Dough Calc app today 👇
