In recent years, artificial intelligence has moved from buzzword to practical tool, reshaping the way we learn and study. To better understand these dynamics, I spoke with two people directly responsible for shaping TripleTen – one of the largest EdTech platforms on the market: Maksim Kulichenko, who leads engineering initiatives, and Anastasiia Prokopenko, responsible for product and student experience.
TripleTen is a bootcamp best known for its innovative mobile learning platform and real-world approach, and was recognized in 2023 for having the top mobile app in the sector. Through conversations with the leading experts I wanted to learn what happens behind the scenes as companies like TripleTen bring AI into the learning process and how business goals, technology, and student outcomes intersect in practice.
The Spark of Innovation: Balancing Vision with Reality
In the tech community, Maksim Kulichenko is widely recognized as one of the key innovators in mobile applications for education and telephony. He was the first in Russia to adopt and champion Flutter at scale, proving it to be the most efficient framework for building mobile products, long before it became an industry trend.
The journey to integrate AI in EdTech is a new challenge – one shaped by both bold vision of innovation and internal skepticism.
What does it mean to you to create technology that helps people learn, rather than just developing a product?
Maksim: For me, it’s about respecting that people’s attention is their most valuable resource. Even when building a marketing-focused or trial application, we have to ensure that users who come to us receive genuine value. They should leave with a feeling they spent their time well, even if they don’t enroll in a full program. My goal is to use technology to help them better understand the profession they’re interested in to bring sharp minds to digital technology.
Was this drive for innovation based on data or intuition?
Maksim: With the mobile app, it was a mix of intuition and past experience (ed. Previous projects in bigtech Yandex, a Russian multinational technology company that operates a popular search engine, main competitor to Google in Russia). I saw the growing trend of mobile-first platforms; people often have a powerful smartphone but no laptop. In a previous project, we underestimated the need for a mobile app. To provide for our users I then had to innovate to build a first ever production Voice over Internet Protocol application in Flutter fast, so we don’t have to write separate code for 2 platforms. It was a success and so this time I thought about it right away. However, most educational bootcamps weren’t using mobile as a primary learning medium so we had to take a bold step here. This was our innovative edge.
Anastasiia: The process of innovation starts with a product hypothesis though. We conduct discovery – analyzing quantitative metrics, qualitative feedback, competitor actions, and academic research. The product manager and development team then brainstorm together. It’s a synergy. The product side brings user needs, while the development side determines how we can build it effectively and beautifully. For example, when we were building the first dashboard system at TripleTen, I was responsible for the product side of the work i.e. defining the requirements from the product team, clarifying what exactly we needed to see in the dashboards, and how we would use this information for decision-making.
At the same time, the engineering team was responsible for determining how to log the necessary data on the platform, where to store it, and how much data we could realistically keep. We worked through these decisions together, because they had the full understanding of the platform’s architecture and the technical implications of each choice. Our analysts, in turn, were responsible for the visualization layer, namely, translating the logged data into dashboards that were both clear and actionable.
So it was a true cross-functional synergy: product defined the “why” and “what,” engineering ensured the “how,” and analytics delivered the “so what.” I’m proud to mention, we have a culture where people genuinely care, and for strategically important projects, discussions can be lively. It often takes several iterations to reach a consensus.
Can you give a concrete example of how this iterative, cross-functional process works in practice, especially when you have to make decisions without clear prior data?
For example, right now, we’re running a series of experiments aimed at improving our user activation metric – the share of learners who complete their first month successfully. Since we don’t have strong prior insights, we’re intentionally testing very different types of hypotheses: for example, whether our students respond more to financial incentives, stricter deadlines, positive reinforcement, or more personalized support.
These hypotheses come from different angles, and different team members naturally believe in different approaches. So the discussions around which experiments to run and how to prioritize them are often very lively but that’s what makes the process strong.
There’s also a lot of hope placed on AI here, because it can help us surface patterns we wouldn’t otherwise see, cut through bias-driven debates, and quickly validate which direction is actually worth pursuing.
The Human-AI Dilemma: Efficiency vs. Empathy
As AI’s role expanded, it brought to the surface a core tension within the company: how to remain a data-driven yet human-centered organization and industry leader.
Integrating AI, what are the biggest concerns or challenges you have to navigate as a team?
Anastasiia: The biggest challenge, which we are navigating, is maintaining the balance between being data-driven and human-centric. We’ve always prided ourselves on that balance. But as AI developed, it became harder.
Maksim: We’re actually going through this challenge right now. We’ve introduced our first AI assistant, and we’re working together to define what should remain in the hands of a human and what can be delegated to AI. I’ve already experimented with how we can innovate with AI to help our students decide which program to attend by building an AI consultant service directly in our mobile app.
Anastasiia: But when a question is connected to motivation, when a student is feeling overwhelmed or discouraged, or when someone is struggling with a concept and needs deeper support, we always direct them to a human. This ensures a more meaningful interaction where it’s needed most. In this way, we preserved the human touch where it truly matters and at the same time upskilled our team to work effectively with our first AI assistant.
However, the rise of AI also means that some support team roles change or become obsolete. It creates a dilemma: we want to invest in these beneficial technologies, but it’s getting harder to define the boundary where we remain truly human-centric. This is an ongoing debate, especially for managers of support teams
Maksim: Frankly, we’re still navigating a major transition. The team doesn’t yet share a unified view on AI, and the topic sparks plenty of debate, including among leadership.
Anastasiia: But even without a top-down decision, the concerns are already there on the ground level. I am leading the support team, and we are discussing what happens to our roles with the rise of the AI bots. There are fears about job security and how our work will change. So, these conversations are happening at a team level, driven by the financial and business metrics we still have to meet.
The Future of Learning: Personalization and Empowerment
Despite the challenges, both leaders are optimistic about AI’s potential to democratize and elevate education.
Looking forward, how do you see AI changing the landscape of education?
Maksim: AI will absolutely change how we learn. We are only seeing the beginning. Education will become far more personalized. We’re moving toward a future where every student can have a personal AI tutor. Research, like Bloom’s 2 Sigma Problem, shows that personal tutoring can turn an average student into a top performer. Secondly, learning will become faster. As market demands change at an accelerating pace, people will need to retrain more quickly, and AI can facilitate that. From my own personal experience, I know how valuable a tutor can be. I had one for just three months at school and achieved very high marks on a final math exam. AI has the potential to give everyone that opportunity, which is super inspiring because we are changing the lives of tens of thousands of users.
Anastasiia: This fits perfectly with our product’s core principles: being accessible and affordable. We’ve always been one of the most accessible options on the market. Our promise is to teach you the skills you need, regardless of your starting point. This implies personalization—some students need more time, others need different tasks. AI makes delivering on that promise more realistic. It can provide the quick, tailored help a student needs, exactly when they need it, reinforcing our competitive advantage of the top-tier EdTech. And this is something I recognized early on. Back in 2021, when I was organizing the EdCrunch conference and selecting topics and speakers, I chose to include a session on LLMs and personalization in education as one of the main topics. I invited the CEO of TripleTen as the speaker because even then it was clear that TripleTen would become one of the leaders in pushing this type of personalization forward – and that this direction would only continue to grow.
with fears of job displacement, how do you help your teams and students stay useful?
Anastasiia: By reinforcing the value of the human touch. The role of an educator or support person won’t disappear, but it will change dramatically. My leadership approach is to be transparent and manage the pivotal process smoothly. Historically, student-facing teams are the ones most affected by every change in the company. For example, when we implemented the flexible learning model and completely rebuilt our support system, or when we started running activation experiments, my team had to adjust workflows immediately. It always needs to stay aligned with every product update. And now, with AI increasingly influencing day-to-day processes, this becomes even more critical. That’s why it’s important for me to manage this pivotal process and ensure that the transition for student-facing teams is smooth, coordinated, and sustainable.
I also focus on individual growth. I talk with my team members about their career goals and find projects that align with their talents and interests, whether it’s in analytics, design, or management. It’s about helping them rotate and find new, interesting roles within the company. That’s how we manage to have the best engineers, managers and educators of the industry.
Maksim: And for me it’s about the safe space to feel confident in discovery. I’ve mentored students for years, and the biggest barrier is often their fear of looking stupid when asking questions. No matter, team, students or users of our platform, we’re already seeing that people are more willing to ask AI because there’s no fear of judgment. This is a huge potential for learning. In the long term, looking 5 or 10 years ahead, it’s clear that AI is progressing exponentially faster than humans can adapt. We will need to redefine our own human value in the labor market. But since the future is hard to predict, we focus on what we can do now to make human resources management and learning strategies better.
And apart from creating a safe environment, I expect my team to take on the user perspective. This is what I also demand from myself: Building Telephony at Yandex, I routed all my personal calls through a prototype app to uncover hidden bugs. Being able to adjust your view from a developer to a user helps to stay curious and create products that actually work
Looking Ahead
As our conversation wrapped up, one thing became clear: AI in education isn’t a clean, linear story. It’s a mix of promising breakthroughs, practical constraints, and very human questions about responsibility, motivation, and the future of work. No one has all the answers, and perhaps that’s why this period feels so pivotal.
TripleTen’s experience shows that the real work rarely happens in grand strategic moments, it happens in the daily debates about experiments, in conversations inside support teams, in decisions about what should stay human, and what can safely move to AI. The future of EdTech is being shaped by people learning how to adapt, collaborate, and rethink old assumptions.
What comes next? More experimentation, more iteration, and more conversations like this one. AI will undoubtedly play a larger role in how we learn — helping personalize education, reduce friction, and make skills more accessible. But the direction it takes will depend on how teams like Maksim’s and Anastasiia’s balance efficiency with empathy, and innovation with the real needs of learners.
If there’s one takeaway from this discussion, it’s that progress in EdTech isn’t driven by technology alone. It’s driven by people who are willing to question their own ideas, adjust when something doesn’t work, and stay close to the human experience at the center of learning.
