Two 17-Year-Olds Kept Launching Apps Until One Took Off

Next month, one of the speakers at TechCrunch Disrupt, one of the biggest startup conferences in the world, will be a college student who started his company in high school.

His name is Zach Yadegari. With his friend Henry Langmack, he co-founded an app called Cal AI when they were both 17. TechCrunch announced on September 24 that Yadegari will speak at Disrupt in October about how the app grew so fast.

The numbers are big. Cal AI passed 15 million downloads and more than $30 million in yearly revenue in under two years. In a deal that closed in December 2025, MyFitnessPal, a much larger company, bought it. Yadegari still runs the app, now as part of MyFitnessPal, while attending college.

But the most useful part of the story for young builders is not the money. It is what happened before Cal AI: a lot of apps that did not take off.

Yadegari and Langmack met at a coding camp when they were 10 years old, according to Forbes. They started building things together right away.

Yadegari's earlier projects included a website that let students play games during school and an alarm clock app that played motivational sound clips. At 16, he sold a gaming website called Totally Science for a six figure amount. It had drawn more than 5 million users during the pandemic.

The pair had a plan. As Yadegari told Forbes, they would keep releasing new apps until one stood out, then go all in on that one.

That plan is worth reading twice. They did not wait for one perfect idea. They treated each app as an experiment. Most experiments are small and quick, and most do not work. The point is to learn from each one and to be ready when one does.

Cal AI is an app for adults who want to track what they eat. You take a photo of a meal, and the app estimates the nutrition in it using image recognition and the phone's depth sensor.

Here is an important note for families. Counting calories is not something kids or teens should be doing unless a doctor or dietitian is guiding them, and it can be harmful for some people. We are not recommending the app. We are looking at how two teenagers built a business, which is a separate lesson from what the business sells.

With that said, the product idea teaches something useful. Photo based food logging already existed in other apps. Cal AI did not invent the concept. It made the experience faster and simpler for people who wanted speed over perfect detail. MyFitnessPal's CEO told TechCrunch that his company's app is for people who want precision, while Cal AI serves people who want it quick. Two products, two different kinds of customers.

Yadegari told Forbes that the product was built around marketing. The design had to be so simple that someone could understand it within three seconds of seeing it in a short video.

Three seconds is not long. Try this test with anything you make:

1. Show a friend your app, drawing, product or poster for three seconds.
2. Take it away.
3. Ask them: "What does it do, and who is it for?"

If they cannot answer, the idea may be fine, but the presentation needs work. Cut words. Make one thing big. Remove anything that is not the main point.

This works for lemonade stands too. A sign that says "Cold lemonade, $1" beats a sign with five flavors, a paragraph about your recipe and a drawing of a lemon wearing sunglasses.

The "keep launching until one works" approach has a hidden rule: each launch has to be cheap. If every idea takes a year and all your savings, you only get a few tries. If each takes a weekend, you get dozens.

Here is how a young builder can run small experiments:

• Pick a tiny version. Instead of a whole app, make one screen, one page or one product.
• Set a deadline. Give each experiment one or two weeks.
• Decide what success looks like before you start. For example, "10 classmates try it" or "5 people ask for another one."
• Look at the results with clear eyes. If it did not hit the goal, write down one thing you learned and move on.

Keep a notebook of every experiment: the idea, the goal, the result and the lesson. After ten tries, that notebook is worth more than any single idea, because it shows you what people actually want.

When one experiment starts working, the lessons on Foundra Kids can help you turn it into a simple plan.

When MyFitnessPal's CEO first met Yadegari, he noticed something small. Cal AI's team held its weekly check in meeting on Sunday nights. Since the founders were still in school, Yadegari worked on the company on weekends, and his team joined him for that meeting.

"This is someone who's not doing this as a hobby," the CEO told TechCrunch. Small, consistent habits like that were part of why the bigger company wanted to buy Cal AI and keep the team.

For a young builder, taking it seriously does not mean giving up school, sleep or friends. It can look like:

• A regular time each week for your project, even just one hour.
• A simple list of what you did last week and what you will do next.
• Following up when you say you will.

Grown ups notice reliability. It is often what separates a young person with a cool idea from a young person people want to work with.

Here is a surprising part of the story. Yadegari told Forbes that building a calorie app was not his "north star." He sold partly because he wanted to move on to building something more impactful.

He also changed his mind about college. At first he did not plan to go. After a summer around other young founders, he decided that a degree would keep more doors open. Now he runs Cal AI while taking classes.

Two lessons hide here. First, your first business does not have to be your life's work. It can be practice for the next thing. Second, big decisions like school are allowed to change as you learn more. That is not failing. That is updating.

Want to try the Cal AI approach at your own scale? Here is a two day challenge.

Saturday morning: Write down five problems you or your friends have. Small ones are best: forgetting homework, losing earbuds, never knowing what to cook for a snack.

Saturday afternoon: Pick one. Build the simplest possible fix. It could be a checklist, a sign, a paper prototype or a one page website made with a parent's help.

Sunday: Show it to five people using the three second test. Ask what they understood and whether they would use it.

Sunday evening: Write your experiment notebook entry. Decide: try again with changes, or pick a new problem next weekend.

Do this four weekends in a row and you will have run more real experiments than most adults ever do.

How old were the Cal AI founders when they started? Zach Yadegari and Henry Langmack were 17 and still in high school when they co-founded Cal AI. They met at coding camp at age 10.

How much did MyFitnessPal pay for Cal AI? The price was not disclosed. Cal AI had more than $30 million in yearly revenue before the deal and projected about $50 million for 2026.

Should kids use calorie tracking apps? Not without guidance from a doctor or dietitian. This lesson is about how the founders built a business, not a recommendation to use the product.

Do I need to know how to code to start? No. Coding helped these founders, but many first businesses are services or simple products. Start with the problem, then learn the skills you need.

What is TechCrunch Disrupt? A large startup conference in San Francisco where founders, investors and tech leaders meet. This year it runs October 13 to 15.

1. TechCrunch Disrupt 2026: Cal AI's Zach Yadegari on how to create viral growth and capitalize on it, TechCrunch, September 24, 2026
2. MyFitnessPal has acquired Cal AI, the viral calorie app built by teens, TechCrunch, March 2, 2026
3. This U30 Kept Launching Apps Until One Worked. Then Sold It To MyFitnessPal, Forbes
4. MyFitnessPal Buys Teens' Calorie App After 15M Downloads in 2 Years, eWeek