Inspiration
Every day millions of people save recipes from the internet and cookbooks. But inspiration rarely becomes dinner. I agree with Eitan Bernath on this.
I noticed a gap between seeing a recipe and actually cooking it. People forget ingredients, lose the video, or feel overwhelmed once they reach the kitchen.
I wanted to build an AI cooking companion that removes friction between discovery, shopping, and execution.
What it does
Make It helps users turn food ideas into real meals. The application can:
-Extract recipes and ingredient lists from a YouTube link -Detect ingredients from a photo of your fridge -Capture ingredients just by copying and pasting text from any source -Generate smart grocery lists based on what you’re missing -Guide you step-by-step with a voice assistant -Let you cook hands-free in an immersive mode -Instead of collecting recipes, people finally cook them.
How we built it
I built Make It using React Native with Expo to move fast while delivering a polished iOS experience. AI is used in multiple layers:
-parsing recipe data from URLs and text -ingredient normalization and matching -fridge image understanding -voice-driven navigation during cooking See video for more details.
Challenges we ran into
Making the Google Voice API to work properly was one of the biggest challenges. It costed a lots of tokens! The biggest challenges were around reliability and clarity. Also I learned that recipes online are messy. Ingredients appear in different formats, measurements vary, and matching them to what someone already has is complex. I also had to design a voice experience that worked in a noisy kitchen while keeping the interface readable from a distance. Finally, building a premium feel in a short timeframe pushed me to carefully prioritize which interactions mattered most.
Accomplishments that we're proud of
I am proud that Make It feels like a real consumer product, not just a demo.
Highlights: Smooth flow from inspiration → grocery → cooking -A cinematic, voice-first step mode -Smart missing-ingredient detection -A scalable subscription model powered by RevenueCat -A design system consistent across all screens -Most importantly, people can complete the full journey from idea to plate.
What we learned
We learned that convenience is king in the kitchen.
What's next for Make It
I see Make It becoming the operating system for home cooking.
Next steps include: -deeper personalization from dietary profiles -smarter pantry tracking -real-time substitutions -creator partnerships -collaborative family shopping
Vision: if you can imagine the meal, you should be able to cook it.
Built With
- ai
- analysis
- and
- backend
- blur
- build
- data
- eas
- entitlements
- experience
- expo.io
- extraction
- firebase
- firestore
- for
- fridge
- frontend
- guidance
- haptics
- icons
- infrastructure
- ingredient
- intelligence
- language
- large
- matching
- mobile
- models
- monetization
- native
- natural
- node.js
- normalization
- processing
- react
- reanimated
- recipe
- recognition
- revenuecat
- speech
- submit
- text-to-speech
- tooling
- typescript
- ui
- vector
- vision
- voice


Log in or sign up for Devpost to join the conversation.