nyam nyam

Inspiration

I was thinking about making app for tracking calories with AI when I found out about this challenge and Eitan's idea about recipe app. I thought that these two concepts overlap and I can learn a lot for my nutrition app.

I did some research and found out that there are tons of apps that extract recipes from videos with AI, but the actual process of extracting was quite cumbersome. That's where I thought that I can improve this process and make importing recipes into app seamless and frictionless.

What it does

App has basically two types of usage. I think of separation in this way:

1. frictionless recipe import - me and probably 9/10 of other people save recipes when we don't actually want to cook them. We save recipes when we're waiting in a queue, sitting on the bus or doom scrolling on the couch.

That's why I wanted users to be able to import a recipe into the app just by sharing it via the share sheet without even leaving the app that they are in.

1. your collection of recipes - once user want to actually cook something he needs to have really good overview of all recipes with sections, search and filtering. That's why there are plenty of sections (by cuisine, quick meals, for lunch, etc.) in this app so that user doesn't have to think and can just pick.

And on top that app supports custom user collections, meal plans and cook mode where it shows recipe steps in large font.

How we built it

I'm senior iOS dev so I know how to make good iOS apps, but honestly I haven't been loving ecosystem lately. In the same time I found out that you can now build apps with native elements in Expo (React Native) so I've decided to try it out (not really responsible decision, but hey I love to learn new things).

So I've ended up spinning multiple Claude Code sessions, fixing multiple bugs, learning A LOT of new things and falling in love with this new ecosystem.

Challenges we ran into

There were tons of them, but probably the hardest one and most important one was recipe extraction pipeline.

I've spend a lot of time testing what tools give best results and to my surprise I've ended up with multiple calls to Gemini (I've tried to combine it with multiple recipe / nutrition databases, but Gemini had better results in the end).

Accomplishments that we're proud of

I'm really proud of native iOS look and how app works, even though it needs to be polished here and there. I'm also super glad that I went for Expo + React Native, because app needs only few modifications and its ready to be released on Android too.

What we learned

I've learned A TON of new things about Expo & React Native and honestly even thing or two about RevenueCat implementation there.

I've also learned that AI is now pretty accurate about specific problems (talking about Gemini).

What's next for nyam nyam

There's still a lot of space for improvement and I have many ideas for features, but first thing is to fill up AppStore data, test app few times, fix last bugs and ship it. jumping happily

Built With

• clerk
• convex
• expo.io
• gemini
• react-native
• revenuecat
• swift
• typescript