Inspiration
Prepist was born from a simple frustration. I was constantly discovering incredible recipes on TikTok and Instagram but actually cooking them was chaotic. My camera roll was full of screenshots, bookmarks were scattered, and ingredients were never structured or easy to follow.
The problem was not discovery. It was execution.
I wanted a way to transform inspiration into something structured, reliable, and cook-ready. That became Prepist.
What it does
Prepist imports recipes from videos, websites, photos, and notes, wherever a recipe exists, we can extract it and convert it into clean, structured, step-by-step instructions.
Supported sources include:
- TikTok, Instagram, YouTube, Pinterest, and Facebook
- Recipe websites and blogs
- Photos of printed cookbooks
- Handwritten notes
- Pictures of meals
Every import is automatically transformed into organized ingredients, clear steps, and a usable grocery list. Beyond extraction, Prepist is designed around inspiration and organization.
The UI is inspired by Pinterest, using a staggered grid layout to make browsing recipes feel visual and intuitive. Users can personalize each recipe by choosing a cover photo from Unsplash, taking a new photo, or selecting one from their camera roll.
Recipes can be grouped into cookbooks for easy organization, and users can build and edit recipes to keep everything structured and clean. Prepist combines powerful extraction with a thoughtfully designed interface that makes saving, organizing, and planning meals effortless.
In short: Inspiration → Structured Recipe → Meal Plan → Dinner
How we built it
Prepist is built with:
- Laravel for APIs, data modeling, and background job orchestration
- React Native with Expo for the mobile app
- Queue workers for asynchronous recipe extraction
- Cloud storage for media handling
- AI-powered extraction to convert unstructured content into structured recipe data
The core system transforms messy inputs into normalized, validated recipe schemas:
Video or Web Content → AI Extraction → Validation Layer → Normalized Recipe Schema
We invested heavily in schema design to ensure ingredients, quantities, and steps are consistent and reliable across imports. Performance optimization was critical. Extraction runs asynchronously so users are not blocked, while infrastructure is carefully tuned to balance responsiveness and cost.
Challenges we ran into
Extraction reliability
Cooking videos are inconsistent. Ingredients may appear in captions, voiceovers, or on-screen text. We iterated extensively on prompts, validation rules, and post-processing to improve accuracy.
Speed vs infrastructure cost
AI processing introduces both latency and cost. We had to carefully balance user experience speed with backend efficiency and queue scaling.
User Experience Speed ↔ Infrastructure Cost
Scope discipline
It was tempting to add editing tools, social features, and advanced meal planning. We deliberately narrowed the MVP to focus on the core value: effortless import and organization.
UX simplicity
Designing something powerful yet intuitive required continuous iteration. We refined flows to reduce friction and keep the cooking experience distraction-free.
Accomplishments that we're proud of
- Building a working recipe extraction pipeline
- Designing a structured, normalized recipe schema
- Delivering a fast, intuitive import experience
- Turning chaotic content into reliable, cook-ready recipes
- Creating a thoughtful, frictionless user experience that makes saving, organizing, and editing recipes feel simple and natural
What we learned
- Structure creates confidence. Clean formatting changes the cooking experience dramatically.
- Speed drives adoption. Users will not wait for complexity.
- Simplicity is a competitive advantage.
- The real problem was not inspiration. It was clarity.
Clarity > Features
What's next for Prepist: Save & Plan Recipes
Next, we are focused on:
- Improving extraction accuracy and performance
- Enhancing grocery list intelligence
- Introducing lightweight meal planning
- Expanding supported content sources
- Continuing to refine UX for clarity and focus
Our vision is to make Prepist the simplest way to go from discovering a recipe to confidently cooking it.
From inspiration to dinner, without the chaos.
Built With
- background-workers
- cloud-hosting
- cloud-storage
- expo.io
- laravel-(php)
- laravel-queues
- llm-apis
- postgresql
- react-native
- redis
- unsplash
Log in or sign up for Devpost to join the conversation.