CookThis!

Stop saving recipes you never make. CookThis turns any recipe video/link into weekly meal plans with smart grocery lists and beautiful step-by-step cooking guides.

Comment

About CookThis

What Inspired This Project

I'm on a mission this year to launch 5 consumer AI apps and transition from my day job as a Design Engineer in Healthcare SaaS to full-time self-employment at my studio, HappyFace Studio. This hackathon felt like the perfect catalyst to launch the first one --> CookThis.

The idea came from my own frustration. I have recipes saved everywhere: Instagram posts, TikTok videos, browser bookmarks across three different devices. When dinner time hits, I'm overwhelmed by choice and can't remember what I even saved or why. And once I finally decide on something? Now I have to figure out what ingredients I need, what I already have, and somehow turn that into a coherent grocery trip. It's exhausting. Most nights, I just order takeout.

This isn't a "me" problem. Everyone I know has 200+ saved recipes they've never touched. The gap between inspiration and execution is massive, and nobody's really solved it.

How I Built It

I built CookThis as a native iOS app using Swift, leveraging Apify to scrape recipe metadata from any URL or video link—TikToks, Instagram Reels, YouTube, food blogs, you name it. Then AI (Gemini Flash 3,0) extracts ingredients, steps, timing, and even cooking techniques from unstructured content. Lastly, an image model through Fal.ai to create beautiful imagery and visual explanations of each cooking step.

The core features:

  • Smart recipe capture that works across all platforms
  • Weekly meal planning with a swipeable interface to curate what you actually want to cook
  • Intelligent grocery lists organized by store sections (produce, dairy, pantry, etc.)
  • Visual cooking mode with hands-free step-by-step instructions and one-tap timers

What I Learned

Building for real people (not enterprise users) is completely different. In Healthcare SaaS, you're solving complex workflows for power users. Here, I'm designing for someone who's tired, hungry, and just wants to know what's for dinner. That constraint forced me to be ruthless about simplicity.

I also learned how powerful AI can be when applied to everyday annoyances. Recipe parsing isn't sexy, but when it works seamlessly, it removes a genuine pain point. The technical challenge was making the AI feel invisible. Users shouldn't think about "AI extracting ingredients," they should just think "wow, that was easy."

Challenges I Faced

The biggest challenge was scope. I wanted to build meal planning, pantry inventory tracking, nutritional analysis, social sharing, recipe reminders... I had to kill features I loved to ship something people could actually use in the hackathon timeline.

Technically, parsing video content for recipes is messy, transcription quality varies, cooking videos often don't follow a linear structure, and ingredient quantities can be mentioned visually or verbally. I had to build robust fallbacks and smart defaults.

This was my first time building something I plan to actually launch and monetize. I am super excited to finalize CookThis and launch it on the 1st of March on the AppStore

What's Next

CookThis is just the beginning. It's the first of five apps I'm launching this year as I build HappyFace Studio into a portfolio of delightful, AI-powered consumer tools. If this hackathon proves one thing, it's that I'm ready to make the leap.

Built With

Share this project:

Updates