Why I'm Building a Cooking App Nobody Asked For

Every night it's the same ritual. I'm in the kitchen, phone in hand, thumb in the search bar of my bookmarked recipes, trying to remember what I can make.

And most nights I make something from that list. Which means most nights I make the same thing I made last week. Chicken. Bolonhesa. The usual suspects.

I don't hate that, to be clear. During the week I want fast and easy — chicken again is fine. But every so often I want a small spin, something a little different. And on weekends I actually want to explore. Cook something new instead of ordering in.

So I do what everyone does: I Google it.

And then it's a miracle if I can even read the recipe. Buried under ads, under a thousand words about someone's grandmother in Ohio, under an autoplay video I never asked for. By the time I scroll down to the actual instructions, I've half lost the will to cook at all.

Here's the thing I only understood after I started building. My bookmarks are a database with a search bar. That's all they are. And a search bar can only ever hand you back what you already put in. So I search what I know, I cook what I know, and my kitchen quietly shrinks down to a rotation of four meals.

What I actually want is dumb and specific: tell me what I can make with the stuff already in my cabinet. Give me the spin — tonight — without a shopping trip.

A database can't do that. You can't search your way to "surprise me with these six things I already have." Only a chef can. Someone who looks at what's in front of you and improvises.

So I'm building the chef. It's called Panela.

Every recipe app on earth is the same architecture from 2010 — a giant database, a search bar, a mountain of SEO on top. Panela skips the database. The AI is the chef. You tell it what protein you've got, it generates the recipe, checks it against your cabinet, offers substitutes for whatever you're missing, and walks you through it step by step. No ads. No grandmother in Ohio. No life story. Just dinner.

It's Portuguese-first, and that part isn't really about language. I speak English fine — better business English than kitchen English, if I'm honest. It's that recipes written in English are grounded in someone else's context. Their pantry isn't my pantry. Half the time I hit an ingredient we simply don't use here, something I'd need a special trip to a specialty shop to even find. It's my heritage that the English recipes don't even comprehend. I love Portuguese food — nearly all of it, liver dishes aside — and I want to cook it in my own kitchen, in my own context, not translated through someone else's.

Now here's the part my advisors would hate.

I have no revenue. Two years of building, zero euros. I just took a day job as a PM to pay the bills. And the very next thing I start is… a cooking app. For myself. The charter literally says it: Revenue — $0. Personal tool. Founder is the user.

Every voice in startup-land is screaming the same thing. Validate the market. What's your TAM. Build something that makes money, not a kitchen toy for one guy.

I hear it. Here's my answer.

For two years I built for imaginary users — people I invented, whose problems I only guessed at. This time I'm building for the one user I can actually trust: me. I know exactly what I need, because I'm standing in that kitchen every single night. And if I need this, chances are other people do too. So I'll build it, use it, quietly show it to a few people, and let them tell me if it's real. If it is, I'll figure out the money — when it's earned, not before.

And the thing that makes all of this possible is the exact thing that felt like giving up: the day job.

Time is on my side now. That's the whole shift. When there's no money coming in, every project has to become a business by tomorrow or you starve — and that desperation makes you build the wrong things for the wrong reasons. The PM job took that gun away from my head. It didn't kill the dream. It funded the honest version of it. Now I can build slowly. I can build for one. I can wait for the truth instead of forcing it.

(It also costs me nothing to run — free tiers all the way down, a real app for zero a month. In 2026 there's genuinely no excuse not to build exactly what you want.)

I don't know if Panela becomes anything. Maybe it's just the tool that finally gets me out of the chicken-and-bolonhesa loop. Honestly, that alone would be worth it.

But I've got a feeling food is going to be a thread for me. My next idea is also food — a way to find the best places to eat around Portugal without scrolling a map for twenty minutes. Turns out when you stop building for imaginary people and start building for your actual life, the ideas don't dry up. They just get more honest.