Supachef

Welcome Screen
Onboarding Quiz
Onboarding Quiz
Intro to AI sous-chef
Import recipe from URL
Imported recipe
Recipe book
Request meal plan
AI generated meal plan
Edit generated meal plan, move to different day/meal

Inspiration

I'm a techie, full-time software engineer, AI enthusiast, mom, homemaker, and everything in between. While I can cook tasty food, I’ll be honest—cooking has always felt like a chore.

To make things more interesting, my 5-year-old turned out to be just as picky as I am. So yeah, meal planning in our house? Not a walk in the park.

Last year, my parents stayed with us for a whole year, and honestly, it was the break I didn’t know I needed. My mom is an incredible cook. Watching her whip up delicious meals, pair dishes like a pro, and make cooking look effortless was nothing short of inspiring. For her, cooking isn’t a task—it’s an art. We were even packing lunch for work every day, and she never let anything go to waste.

When they left, I found myself falling back into the same routine: bland, repetitive, “meh” meals. I missed the variety. I missed the joy.

That’s when it hit me—if I could bottle up the way my mom planned and cooked meals into a mobile app, maybe cooking could be... not just bearable, but actually enjoyable.

And that’s how SupaChef was born.

What it does

Think of SupaChef as your kitchen sidekick, only smarter—and always within reach on your phone.

This mobile app helps you cook more at home without the stress. No more standing in front of the fridge wondering what to make—SupaChef plans meals that actually fit your life: what’s in your pantry, what your kid will eat, what you feel like eating—and helps make sure you’re not tossing food at the end of the week.

It also lets you organize all those saved and bookmarked recipes from Instagram, YouTube, or wherever you scroll for food inspo—finally, a single place to keep track of them all. You can even add your own recipes, tweak them, and manage your entire personal recipe collection without the chaos of screenshots and browser tabs.

The best part? SupaChef thinks ahead. It helps you reuse ingredients across meals, suggests ideas for school lunches, and turns your weekly meal plan into something that feels a little more intentional… and a lot less painful.

How we built it

I think you’ve got an idea of how wild my daily routine is—now multiply that by 100. Honestly, this idea wouldn’t have seen the light of day if it weren’t for Bolt. It made prototyping fast, fun, and super manageable even with everything else going on.

I also got lucky with an amazing teammate—he had just wrapped up his co-op at my company and was excited to jump in when I told him about this hackathon. 🙌

We kicked things off with ChatGPT for brainstorming, defining the problem, generating the PRD, and even helping create some of the visuals and user flows. Having an AI sounding board at that stage made it so much easier to go from idea to execution.

We used Bolt to build the mobile app, which gave us a fast and flexible way to create a native-like experience without all the heavy lifting. Bolt’s built-in GitHub integration made team collaboration and version control effortless.

For the backend, we chose Supabase for authentication and database storage. We also used Supabase Edge Functions to handle server-side logic for AI calls and meal plan generation—making the app feel fast and responsive without overcomplicating the architecture.

Finally, we integrated the Groq (Mistral) model to power the AI-driven meal planner. That’s where SupaChef really came to life—serving up smart, personalized plans in seconds.

It was a real team effort—with AI, human creativity, and just enough caffeine to keep things rolling. ☕💻

Challenges we ran into

URL parsing from different social media sites turned out to be trickier than we expected. Every platform formats things differently, and there wasn’t a universal solution that just worked. After wrestling with that for a while, we ended up going with the LinkPreview API, which helped us simplify things and move on.

Another big challenge was working with Supabase Edge Functions and integrating the LLM (Groq/Mistral). Figuring out the right prompt, handling API responses gracefully, and making sure everything played nicely with the app took a good chunk of our time. It was one of those “everything works in isolation but breaks when connected” moments—but we learned a ton along the way.

Accomplishments that we're proud of

We built an AI-powered app. Like... really built it. It's almost a mini-RAG system (Retrieval-Augmented Generation), and we pulled it off without losing sleep over things like getting every pixel just right.

What makes us especially proud is that we didn’t just build something for the sake of using AI—we built something practical. Something that can actually help people make smarter, healthier choices without the stress that usually comes with meal planning.

It started as a personal pain point, but seeing it take shape into an app that could help others like me? That’s the real win. 💪🍽️

What we learned

We definitely got better at prompting. Like, a lot better. Using Discuss Mode in Bolt was a game changer—I could break things down into subtasks, revise prompts, and fine-tune until I was happy with the outcome. Honestly, 8 times out of 10, I didn’t even have to revisit it. That’s a win in my book.

I also learned that working with AI requires a different mindset. You can’t just throw everything at it and expect magic. Doing things step-by-step, in a clear and logical flow, made a big difference. Trying to do too much at once? Been there—didn’t work.

There were maybe one or two moments of “ugh, why is this not working?!” frustration, but overall, the combo of Bolt’s incredible tooling and a layered, intentional approach to prompting made the experience smooth and surprisingly fun.

What's next for SupaChef

We’ve got big plans for where SupaChef goes next:

Make it a full RAG app
Incorporate vector search to bring smarter, more context-aware meal suggestions.
Add a Pantry feature
Automatically recognize staple ingredients like rice, flour, and spices—so the app can suggest meals based on what you actually have, not just guess.
Enable “Plan Ahead, Shop Later”
Generate a smart grocery list based on your meal plan, so you can shop intentionally and avoid overbuying.
Make it more health-focused
Track calories, macros, and dietary goals, and get meal plans tailored to your personal health journey.
Support even more preferences
Expand personalization options to better reflect real-life needs, cultural foods, and family dynamics.

This may have started as something to help me, but it’s grown into something with the potential to help so many others—and I’m excited to keep building!