• 

  The landing page, fully designed and built by Bolt.

• 

  Onboarding is a breeze, and Bolt determined the steps required.

• 

  A simple input field for tracking calories of any type of meal.

• 

  Mouthful accurately estimates the calories and macros of your entire input.

• 

  Keep a historical overview of your tracked calories and meals.

The Story Behind Mouthful

From Dead Ends to a Spark

I’ve wanted to participate in the Bolt Hackathon since the day it was announced. Over the weeks I explored a bunch of ideas. Puzzles, games, photography tools, productivity apps. But none of them felt right. Nothing felt worth building. Nothing felt genuinely useful.

The Frustration That Sparked It

Then came the final weekend of the hackathon.

I had decided to start tracking my calories again to support my strength training routine. But the moment I opened my usual food tracking app, I felt the same dread I always do.

Search, scan, estimate, log. Do that for every single meal. It’s a tedious loop that turns a simple habit into a chore.

That’s when the idea hit me. What if I could just describe what I ate in plain language, and an AI worked out the calories for me?

That was the moment Mouthful was born.

What I Built

Over the weekend I built a working MVP of Mouthful. It’s an AI calorie tracking app that lets you:

• Type what you ate in natural language
  
• Get an instant calorie and macro breakdown
  
• Track against your goal: lose, maintain, or gain

It understands vague portions, and gives you clarity without the manual work.

What I Learned

The biggest lesson I learned is that the best ideas usually come from solving your own pain. No amount of brainstorming helped me land on this concept. It only clicked when I faced the exact problem I needed to solve.

The moment was natural. The need was real. And the solution became obvious.

The Challenges

Natural language is unpredictable. People don’t say “213 grams of chicken breast.” They say “a bit of chicken.” Teaching the AI to interpret those phrases in a useful way was a real challenge, but a worthwhile one.

Another challenge was deciding how much input to require. I didn’t want another boring form. So I focused on speed and smart defaults instead.

The Outcome

Even if I don’t win, I’m proud of what I’ve built. Mouthful is something I’ll actually use every day.

It solves a real problem. It’s not a gimmick. And for the first time during this hackathon, I built something that genuinely matters to me.

Built With

• css
• eslint
• html
• javascript
• lucide-react
• netlify
• openai-api
• postcss
• react
• sql
• supabase
• tailwind-css
• typescript
• vite