Asilbek Ashurov launched a SAT tutoring business when he was 16. Now 19, he is building Makon AI with his co-founders — an AI startup that creates personalized study plans, tracks students’ progress, and helps high school students prepare for exams independently. The platform has more than 50,000 registered users, and the company has already raised $380,000 in funding.
For the joint Digital Business and Astana Hub project “100 Startup Stories of Central Eurasia”, Asilbek shared how the team promoted the product among high school students, why he took a leave of absence from the University of Chicago, and why they chose to focus primarily on the US market.
“Within just a few weeks of launching, we were making around $5,000 a month”
– Tell us a little about yourself. Where did you grow up, and what were you doing before starting your first business?
– I grew up in Gulistan, a small city in Uzbekistan about 120 kilometers from Tashkent. Later, I studied at a Presidential School, one of the strongest schools in the country.
I did well at school and was one of the top students in my class. Many of my classmates, including me, were preparing for the SAT — a standardized test widely used for college admissions in the United States.
By that time, the SAT had relatively recently moved to a digital format.
My co-founder Kamoliddin Yaxshilikov and I saw an opportunity and decided to take advantage of it. Demand for SAT preparation was growing, but there were not enough good materials or clear guidance on how to prepare for the new format. So in the summer of 2024, we launched SAT Makon, offering preparation for the SAT and other international exams.
– How did you find your first students?
– Telegram was our main acquisition channel. I was active in different communities and channels where high school students interacted, and we also had our own groups. That helped us attract our first users quite quickly.
It also helped that we were a young team. We were willing to make bold decisions and could react quickly to changes. We simply saw an opportunity and took advantage of it. Basically, we entered the market at exactly the right time. Within just a few weeks of launching, we were making around $5,000 a month. A year later, we were already competing with companies that had been in the market for three or four years. Today, thousands of students from 24 countries study with SAT Makon.
– You started the business while you were still in high school. What happened with your education after that?
– I enrolled at the University of Chicago and completed my freshman year. But this fall, I took a leave of absence for at least a year. By that point, we had already launched a new project in the US — Makon AI — and I decided to focus on it full-time. I plan to return to university once I feel that the key processes in the company are properly set up.
SAT Makon continues to operate with the team we hired, while we, the co-founders, are now mainly focused on Makon AI because we see much greater potential for growth there.
“If we hadn’t changed anything, we would eventually have needed thousands of teachers”
– Why did you decide to launch a new project if SAT Makon was already working well? What was missing from the previous model?
– At some point, we realized that the previous model was difficult to scale globally. If we hadn’t changed anything, we would eventually have needed hundreds, and possibly thousands, of teachers. And the larger the team became, the harder it would be to maintain the quality of teacher selection and onboarding, the lessons themselves, and the feedback students received.
Makon AI co-founders
With AI developing so quickly, we saw an opportunity to solve this problem with technology. That is how the idea for a separate AI product came about — one that helps students prepare independently, analyzes their progress, and adjusts the study plan to each individual user.
We launched Makon AI in late October or early November 2025.
At the same time, SAT Makon continues to operate as a separate business: we still have our learning centers and online programs. So launching Makon AI was not really a pivot in the traditional sense. It is a separate project that grew out of our experience in education.
– How does the platform work?
– Right now, the platform helps students prepare for the SAT and ACT. First, a user goes through an onboarding process with around 15 questions about their current level, academic goals, preferred learning format, how much time they are ready to spend studying, and other factors.
Based on those answers, the system creates a personalized study plan and distributes assignments across the days the student chooses. Users also have their own analytics dashboard where they can track their results. There is a separate section for reviewing mistakes, and students can get hints and explanations from an AI tutor.
– What can the AI tutor do?
– It can generate video walkthroughs and explanations for specific SAT questions. It also supports students while they are solving problems: the system knows the context of every question on the platform and can immediately provide a hint or explanation.
– What is actually “under the hood” of Makon AI? Do you use existing models or your own?
– We use existing models through APIs and build our own system on top of them.
Our main technology is the personalization layer. We collect data on a student’s current level, goals, schedule, results, and mistakes. Based on that, our algorithm builds a study plan and continuously adjusts it as the student progresses. For example, if a student keeps making mistakes on a particular type of question, the algorithm can devote more attention to that topic.
“On average, students improve their SAT scores by around 150 points within three months”
– How do you reduce the risk of the AI giving an incorrect answer or explanation?
– We have tested the system many times and are seeing very high accuracy, so the probability of it giving an incorrect answer or explanation is quite low.
– You have around 5,000 questions in your database. Where do they come from?
– They come from different sources. Some are official questions released by the organizations that administer the exams and allow them to be used. The rest are created by our team.
– How effective is Makon AI? What results are students seeing?
– Based on our data, students improve their SAT scores by around 150 points on average within three months. We calculate this by comparing their scores before they start preparing with Makon AI and after.
– Where are your students from, and how many users do you have?
– They are mainly high school students from the US. We have more than 50,000 registered users in total, around 500 of whom are paying users.
– Why did the US become the main market for Makon AI?
– We originally built the product around SAT and ACT preparation, so the US was the most natural market for us. That is where the main audience for these exams is, and the potential user base is much larger. So from the beginning, we have been building the product primarily for American high school students.
At the same time, we have maintained our ties with Uzbekistan and Central Asia. For example, on June 30, 2026, Uzbekistan’s Youth Day, I presented Makon AI to President of Uzbekistan Shavkat Mirziyoyev at an exhibition of youth projects.
Photo: Press Service of the President of Uzbekistan
“Within the first two months, MRR reached around $12,000”
– Who was on the team from the beginning, and how many people are working with you now?
– When we launched the first business, I had one co-founder, Kamoliddin Yaxshilikov. By that point, he had already been working in SAT preparation for more than four years, including as a teacher. It was important for us to have someone on the team who understood the market and the needs of students really well.
When we started building the tech startup, we needed strong technical expertise. That is when Asadbek Noyibjonov joined the team as co-founder and CTO of Makon AI.
Our team is small, so the three of us do most of the work ourselves: product, marketing, and operations. We work on Makon AI full-time — I would even say more than full-time. For specific projects, we sometimes bring in contractors for a few months.
We also work with experts who have achieved top scores on these exams themselves and have years of experience preparing students for them.
– How does Makon AI make money?
– We use a subscription model. Depending on the country, the subscription costs between $39 and $59 per month.
– When did the startup first start generating revenue?
– We started seeing recurring revenue about 10 days after the launch.
At first, we shared the product with people in our network – friends and family – and then started taking marketing more seriously. We ran UGC campaigns and experimented with different growth strategies.
The project grew very quickly in the beginning. Within the first two months, MRR reached around $12,000.
– As I understand it, you are also developing a B2B business?
– Yes, we have started working with schools in the US. We currently have two active B2B partnerships, and a few more educational organizations are in the onboarding process.
It is still a relatively small part of the business. The contracts are in the mid-four figures per month. At this stage, we are mostly learning how to work with this segment – seeing how schools use the product, what problems they run into, and collecting feedback.
“We spent several months in Palo Alto and started thinking on a much bigger scale”
– How much money did you invest at the beginning?
– At the very beginning, when we launched SAT Makon, we did not need much money. Our main investment was our time.
– What about funding for Makon AI?
– In total, we have raised around $380,000. That includes earlier angel checks and a $300,000 round we closed in June 2026 with IT Park Ventures, Yoshlar Ventures, and AloqaVentures.
– What was the fundraising process like?
– We started reaching out to venture funds in early spring. At the time, I was still at university, and I was handling most of the fundraising myself. We went through due diligence, prepared the necessary documents, answered follow-up questions, and stayed in constant contact with investors. It took us around three months to close the round.
– Besides investment, what other support did you receive to develop the project?
– We took part in Alchemist Silkroad – Silicon Valley Residency, an accelerator program run jointly by Astana Hub, IT Park Uzbekistan, Silkroad Innovation Hub, and Alchemist.
We spent several months in Palo Alto, California, which gave us an opportunity to immerse ourselves more deeply in the US market.
During that time, our perspective changed a lot — we started thinking on a much bigger scale. Probably the most valuable part was being in the Silicon Valley environment itself: being around other founders, investors, and experts.
– Where would you like Makon AI to be in two or three years? What are you planning to focus on?
– Right now, we are focused on building the product for high school students. We want to make the process of preparing for international exams as clear, personalized, and affordable as possible. In the future, we may expand into other areas of education, but that is not our immediate priority.