How I Built Aviation Infinity: Flight School to 50K Students
A student pilot struggling with outdated materials across 33 authorities. So I built the first AI flight academy, now serving 50,000+ students globally.

In 2008, I was sitting in a flight school classroom, staring at a 400-page meteorology textbook that was somehow supposed to prepare me for one of the hardest professional exams in the world.
The Airline Transport Pilot License (ATPL) exam covers 14 subjects. Meteorology. Air law. Navigation. Aircraft systems. Human performance. Flight planning. Mass and balance. Each subject has its own textbook, its own question bank, and its own pass mark. Miss one subject and you retake the entire block.
And here's the part that makes it truly painful: every country does it differently.
EASA (Europe) has one format. FAA (United States) has another. CASA (Australia), DGCA (India), CAAC (China), CAA (UK post-Brexit): 33 aviation authorities worldwide, each with their own syllabus, their own question style, and their own regulatory quirks.
The tools available to study? Printed question banks. Static PDFs. Maybe a CD-ROM if the flight school was progressive. Nothing adaptive. Nothing intelligent. Nothing that knew where you were weak and focused you there.
I was an aeronautical high school student at Giulio Cesare who could code iOS apps. The solution was obvious.
Building the First Version
The first version of Aviation Infinity was embarrassingly simple.
A web app with a question bank, organized by subject and authority. You could take practice exams, see your score, and review wrong answers. That was it. No AI. No adaptive learning. No fancy features.
But it solved a real problem. Student pilots could practice anywhere: on their phone during a bus ride, on a tablet between flights, at home after ground school. The question bank was organized the way real exams were structured, not the way textbooks were written.
I launched it to my own flight school first. Word spread to the next school. Then the next city. Then the next country.
Within months, I had students from Italy, Belgium, the UK, Spain, and Germany. All through word of mouth. Zero marketing budget.
That's when I learned something critical: in niche industries, if your product actually solves the problem, the community does the marketing for you. Pilots talk to other pilots. Flight instructors recommend tools to their students. Flight schools tell other flight schools.
The 33-Authority Problem
The real engineering challenge wasn't the app. It was the content.
Every aviation authority has different rules. EASA uses multiple-choice with four options. FAA uses a different question format. Some authorities weight certain subjects more heavily. Some have oral exams in addition to written tests. Some update their question banks annually. Some haven't updated in a decade.
Building for one authority is a product. Building for 33 authorities is an infrastructure problem.
I had to create a content pipeline that could:
- Ingest questions from different formats and structures
- Map them to a unified taxonomy of topics and subtopics
- Tag them by difficulty, authority, and subject
- Keep them updated as regulations changed
- Generate new questions that matched each authority's style
This is where AI entered the picture, not as a marketing buzzword, but as the only practical way to scale across regulatory frameworks. Manual curation across 33 authorities would require a team of dozens. An AI system that understood aviation taxonomy could do it with one engineer and a handful of subject-matter experts.
The AI Instructor
The biggest upgrade came when I built the adaptive learning engine.
Traditional exam prep is linear. You open a textbook, start at chapter one, and grind through to chapter fourteen. Everyone gets the same material in the same order, regardless of what they already know or where they struggle.
Aviation Infinity's AI instructor works differently:
Diagnostic assessment. Before you study a single page, the system tests you across all subjects to build a baseline map of your knowledge. It knows that you're strong in navigation but weak in meteorology before you've opened a textbook.
Adaptive question selection. Every question you answer updates the model. Get three questions right on cloud formations? The system moves on. Get one wrong on pressure systems? It drills deeper, approaching the concept from different angles.
Authority-aware preparation. The system knows whether you're studying for EASA, FAA, or CASA. It doesn't just change the questions. It adjusts the emphasis, the format, and the regulatory context. An EASA student gets different weightings than an FAA student, because the exams test different things.
Spaced repetition. The system brings back questions you got wrong at scientifically optimized intervals. Not too soon (you'll remember from short-term memory) and not too late (you'll have forgotten completely).
The result: students using the AI instructor pass at a 92% rate. The industry average hovers around 70-75%.
Scaling to 120 Countries
Growth in aviation is different from growth in consumer tech.
You don't scale with viral loops or referral codes. You scale through institutional trust. A flight school has to believe your platform won't teach their students the wrong material. An aviation authority has to trust that your content is compliant. A student has to trust you with the most important exam of their career.
We built that trust through partnerships. Today, Aviation Infinity works with 140+ partner flight schools across 120 countries. Each partnership means the school has vetted our content against their curriculum and their authority's requirements.
Some of these partnerships took months of conversation. Flight school owners are cautious people, responsible for training the next generation of commercial pilots. They don't adopt new tools on a whim.
But once a school is on board, their students follow. And when those students pass at 92%, the school tells every other school they know.
The Electronic Flight Bag
Aviation Infinity outgrew exam prep.
Pilots need tools beyond studying. They need flight planning. Navigation. Weather briefings. Weight and balance calculations. Document management. Logbook tracking.
So I built an Electronic Flight Bag (EFB) into the platform, the same category of tool that airlines use on their iPads in the cockpit, but designed for student and general aviation pilots.
The EFB connects the learning experience to the flying experience. You study air navigation in the classroom, then use the same platform's navigation tools in the cockpit. The theory and the practice live in one place.
This is something consumer tech takes for granted but aviation rarely does. Pilots typically use one app for study, another for flight planning, another for weather, another for logbooks. Every transition between tools is friction. Every separate login is a chance to lose data.
What Building for Aviation Taught Me
Seventeen years of building Aviation Infinity taught me principles that apply far beyond aviation:
Regulated industries reward patience. You can't move fast and break things when pilots depend on your content. Every feature gets tested. Every question gets reviewed. Every regulatory update gets incorporated before it goes live. This is slower than consumer tech. It's also why competitors rarely last, because they don't have the patience for compliance.
Niche markets are bigger than they look. "Student pilots" sounds small. But there are hundreds of thousands of people studying for pilot licenses worldwide at any given moment. And they study intensively for 12-24 months. That's a deeply engaged user base with high lifetime value.
Domain expertise is the ultimate moat. I'm a commercial pilot. I've taken these exams. I've sat in the classrooms. I know what it feels like to open a 400-page meteorology book at 6 AM after a 4-hour flight lesson. No competitor who doesn't fly can replicate that understanding.
The best AI products are invisible. Students don't use Aviation Infinity because it has AI. They use it because it helps them pass. The AI is the engine, not the product. The product is confidence on exam day.
50,000 Students Later
Today, Aviation Infinity is in 120+ countries. 50,000+ students have used it. 140+ flight schools partner with it. The 92% pass rate is the number I'm most proud of, not because it's a good metric, but because it means real people got their pilot licenses who might not have otherwise.
Every few months, I get a message from someone who passed their ATPL and is now flying for an airline. Some of them started studying on Aviation Infinity when they were teenagers. Now they're first officers on A320s and 737s.
That's the thing about building for regulated industries. The stakes are real. The impact is tangible. And if you build something that actually works, it compounds over years in a way that trend-chasing products never do.
Aviation Infinity was my first product. Seventeen years later, it's still growing. And everything I've built since, from ClickAi to every product that followed, follows the same principle: find the industry that technology forgot, and build the tool that professionals deserve.
Enjoyed this article?
I write about building products, AI, aviation, and the journey of entrepreneurship. Follow along for more.
Keep reading

New Pilot Shop & Want To Be a Pilot: Niche Communities
I built an aviation e-commerce store and a pilot mentoring platform for tiny markets. Small communities with deep needs are more valuable than large, shallow ones.

Building for 33 Aviation Authorities Taught Me Regulated AI
Most AI startups avoid regulated industries. Here is what a decade inside aviation, legal, and healthcare taught me about building AI that must be right.

The Aviation Portfolio: How Five Products Serve One Industry
Inside the strategy behind building five complementary products for the aviation industry. Aviation Infinity, Avioyx, AvioSharing, New Pilot Shop, and Want To Be a Pilot.