How I’ve built an AI Meal Planner that saves people 5 hours/week of calorie counting

The Problem

When I started going to the gym, I’ve learned about “calorie counting”. You eat more calories than you need, you gain weight. You eat less calories, you lose weight. Simple enough until you realize that a small chocolate bar can have more calories than an entire lunch.

So people lean often on apps for help: track how much they’ve eaten or to understand nutrition better.

When I did the same, I found a major issue: I was staying all day with the phone in my hand, inputing data into different apps, instead of focusing on my goal.

Why Existing Solutions Weren’t Great

All apps I’ve tried had the same flaw for me:

1. They were fighting for my time - Notification to add eaten food, notification to add water intake, notification to add exercies. They are all data input solutions.
2. Instead of having a partner that guides me, I was figuring it out alone, and the app was helping with calculations, but they weren’t giving me direction.
3. Interfaces were small, complicated and filled with ads

“Why is that? - I asked. AI is getting smarter and smarter. Why isn’t there an app that just tells me what I should eat, instead of me spending time after every meal to check if what I’ve eaten is enough or too much?”

The Vision: What If...

I’ve sat down thinking:

• What if I build an app that asks people, like in therapy, about their current lifestyle, goals, needs
• What if the app instead of asking what food was eaten, just tells people what they should eat
• What if I use AI to give users tailored meals, teach them how to cook, what to buy weekly, fully personalized for everybody

This way:

• They just follow a given path, instead of spending hours inputing food
• They have a reliable source, that always tells them what they should eat, how to cook and what food to buy
• They know their progress and how long until their goal is reached

Building the Solution: From Idea to Reality

Research & Validation

We’ve talked to 30 nutritionists, and what their clients needs were. People needed a way to count calories, no question. The nutritionists job is to do the calculations for people and to recommend healthy foods based on that.

From this finding alone we can build an AI product for 2 categories:

1. For personal trainers to automate their work
2. For people who want to track calories without hiring a personal trainer

I’ve decided to build an app for second category, because I was part of the target audience, so I could test it on myself until it was great. The validation was done with real nutritionists guiding every new feature.

Design Decisions

The whole app is designed to be as simple as possible:

• For busy moms, busy leaders and young people
• The navigation needs to be seamless, not involving thinking
• Avoid adding any learning curve at all if possible
• No ads in free version, just limited taste of the premium tier

Simplicity as a Core Principle

We’ve made an important decision early on: every user interraction should be done by pressing ONE button → “The Next Button”.

Weather it’s saying “Continue” or “Generate” or “Mark as Eaten”, the user should always know to press the button, in the same place, everywhere in the app.

The Results

An app that WON podium at lunch against 35 other products

5 out of 5 stars rating and great reviews

People of age 15 to 57 started using it organically

What are they saying:

Takeaway lessons:

• Don’t settle until you find the right way to satify your audience. This project went from a PDF Generated with AI → Web App → Mobile App.
• Test your product against an expert in the field. If they say it’s good, then an everyday consumer would say it’s great too.
• A quality product lowers the marketing cost long-term as people will talk about it. Word-of-mouth is FREE.