Recipe Keeper & Meal Planner

Inspiration

I spend most of my time working at a computer, and cooking is my way to disconnect and reset. Over time, my recipes ended up scattered across YouTube, TikTok, browser tabs, and notes. Rewriting and manually organising them always felt like unnecessary friction. I genuinely love cooking, which is why I immediately chose Eitan Bernath as the curator - their idea and approach perfectly match how I cook and collect recipes

What it does

Recipe Keeper & Meal Planner helps users save and organise recipes with minimal manual input. Recipes can be created manually, imported from articles and videos directly from the browser or social apps, or recognised and searched by a photo of a dish, handwritten notes, or a text description. All recipes are organised into cookbooks, with support for meal planning, grocery lists, and basic nutrition tracking

Fast recipe and ingredient entry is powered by NLP (Natural Language Processing)

It allows users to enter ingredients and quantities as plain text (for example, “2 eggs” or “200 g flour”), automatically parsing them into structured data without manual field separation (quantity and unit detection, ingredient categorisation, and emoji suggestions).

A dedicated Cooking Screen presents instructions step by step and provides contextual ingredient and measurement tips, so users can interact with their phone as little as possible while cooking

How we built it

The app is built with SwiftUI and designed specifically for iOS 26. NLP (Natural Language Processing) is used to speed up ingredient input by parsing plain text into structured data

A proxy server (backend) is implemented in TypeScript. RevenueCat handles subscriptions, while Supabase powers the server-side infrastructure

The primary development platform is Xcode, with Antigravity and Codex used as supporting tools

Challenges we ran into

The main challenge was simplifying and minimising manual text input. The goal was to achieve high-speed recipe creation and ingredient entry for grocery lists without sacrificing clarity or control.

Another challenge was optimising parsing performance. It needed to be fast and reliable enough to feel almost invisible to the user, without slowing down the core workflow, especially when importing content or adding ingredients on the go

Accomplishments that we’re proud of

• Multiple low-friction ways to create recipes
• Fast ingredient input powered by NLP
• A Cooking Screen optimised for real cooking scenarios: it shows only the ingredients required for the current step and automatically converts measurements to the selected system (useful for users unfamiliar with metric or imperial units)

What we learned

During development, it became clear that speed and minimal interaction matter more than a large number of features. The faster a user can save a recipe or add ingredients, the higher the product’s value in real-world usage

What’s next for Recipe Keeper & Meal Planner

• Deeper integrations with reminders (Apple Reminders, Calendar, Widgets) and shopping services (Instacart, Walmart)
• Cooking Screen improvements: auto-scroll, vertical mode, and more settings to minimise device interaction
• Smart “stacks” in the Grocery List: automatic grouping of identical ingredients from multiple recipes with quantity aggregation
• Analytics and stability improvements with TelemetryDeck and Sentry
• An inspiration screen with curated recipes
• Sharing: recipe sharing and social interactions between users within the app
• Onboarding improvements

Monetization

The app uses a subscription model focused on core value. The free tier allows users to store and view recipes with basic functionality. The subscription unlocks unlimited recipes, article and video import, AI-based photo recognition, and an expanded detailed recipe view.

In the future, integrations with online shopping services may also include affiliate programs

Please Note

The demo video was recorded on a device with a smaller screen (iPhone 13 mini) running iOS 26. Minor performance drops may appear due to screen recording and do not reflect the actual performance of the app

Built With

• antigravity
• codex
• lottie
• revenuecat
• supabase
• swiftui
• xcode