In June this year, the Hong Kong Government unveiled the “Blueprint for Digital Education Development in Primary and Secondary Schools”, accelerating the integration of artificial intelligence into education. Generative AI has rapidly found its way into classrooms and “model answers” become just a click away. As a result, a fundamental question emerges: What is the core value of school education?
Haiyang Xin(Tony), who holds a PhD in Information Engineering from the Chinese University of Hong Kong, founded CocoRobo ten years ago. He has built STEM curricula and AI learning platforms that today serve hundreds of schools across Hong Kong and the Greater Bay Area.
In this issue of CUbicZine, we follow the entrepreneurial journey of this Hong Kong drifter scholar, discover how he has chosen to respond to the relentless tide of technological change—not by moving faster, but by building educational products with greater care, purpose, and responsibility.


Books Sparked the Dream. MIT Turned It Into Action.
Born into a family of educators, Tony grew up surrounded by books. Fascinated by stories of scientists changing the world. During his academic years, he was deeply inspired by tech entrepreneurs such as Steve Jobs and Peter Thiel, shaping a distinctive view of entrepreneurship.
To him, the best entrepreneurs are not merely inventors. They are master orchestrators—people who know how to connect technology, talent, and market needs to create meaningful impact.
A defining turning point came during his PhD studies, when he was awarded funding to undertake an academic exchange at the Massachusetts Institute of Technology (MIT), an experience that prompted him to move beyond theory and into practice.
“No one taught me how to be an entrepreneur,” he recalls. “At MIT, if students had an idea, they would simply go out and make it happen. That, in itself, was a powerful lesson.”
He jokes that the one thing he did not spend much time on during those six months was research. Instead, he found himself captivated by MIT’s vibrant entrepreneurial culture. He sat in on classes whenever he could, immersed himself in startup knowledge, and connected with like-minded founders and entrepreneurial communities. Gradually, he realised that starting a business was far less out of reach than he had once imagined.
This ‘get your hands dirty’ spirit of MIT left a profound impression on him.
“Nothing to lose. Why not take on the hardest challenge while I’m still young and at my most creative and capable?”

After completing his PhD in 2016, he decided to take the leap and set up his own business. Tony’s original ambition was simple: use technology to build something meaningful. His first idea was a consumer robotic dog. To pursue it, he founded the EdTech company CocoRobo. However, the effort was hampered by technological limitations and a market that was not yet ready.

Instead of giving up, he drew inspiration from MIT’s maker culture.
“Why not teach children to build robotic dogs?” He shifted his approach, and the idea arrived at exactly the right moment. Hong Kong was ramping up its STEM education initiatives. Tony pivoted the company toward STEM teaching materials and gradually expanded his offering to include both curriculum and teacher training.

When CUHK Helped Keep the Dream Alive
The transition from student to entrepreneur was far from smooth.
In the early days, Tony relied entirely on personal savings accumulated from scholarships. Within a year, most of that money had been depleted. Despite repeated doubts about the commercial viability of his idea, he was able to persevere thanks to the University’s strong support.
“Without CUHK’s support, I probably would have given up and gone looking for a job.”
CocoRobo received support from CUHK’s entrepreneurship and innovation initiatives, including the Pre-Incubation Centre (Pi Centre) and the Centre for Innovation and Technology (CINTEC). The Technology Start-up Support Scheme for Universities (TSSSU) proved particularly crucial, providing funding at a time when the company faced significant financial pressure.
Many of the company’s earliest milestones—from launching its first product and securing its first school client to organizing its first STEM competition—were achieved with support from the University community.
What Tony values most, however, is not the funding itself, but the culture behind it: an environment that encouraged experimentation, welcomed risk-taking, and accepted failure as part of the learning process.


By visiting schools, discussing partnerships, and conducting classroom demonstrations, and engaging with principals, teachers and educational organizations, Tony gradually established a ‘product–curriculum–competition’ business model.
In 2018, the company secured angel investment and became one of the first startups to enter the InnoHub at the CUHK Shenzhen Research Institute.
Once survival was no longer the primary concern, Tony’s focus shifted from HOW to build a business to WHY it mattered.

“Learning shouldn’t be reduced to examination scores or endless drills. In the real world, problems are solved through an interdisciplinary approach.”
For him, STEM education was about cultivating the problem-solving mindset and collaborative skills. Seeing his startup tackle real societal needs convinced him that he was on the right path.
Returning to the Essence of Education
The pandemic marked another major turning point.
As the company expanded its presence in Shenzhen and generative AI began transforming industries worldwide, it embarked on what Tony describes as a “second startup” journey—reimagining its technology, products, and business model around AI-powered learning systems.
Yet as he worked closely with educators, he discovered that their greatest concern was not learning how to use AI.
It was what AI might do to learning itself.
Teachers worried that students would become overly dependent on technology and lose the habit of independent thinking. Educators themselves could be reduced to operators following AI-generated instructions.

“The most dangerous thing about AI isn’t that it’s not intelligent enough. It’s that it’s too seamless.”
While many companies raced to release increasingly powerful AI products, Tony deliberately slowed down and began asking what learning actually requires.
His search led him deep into the research literature on learning sciences, where he found a recurring insight: meaningful learning is often accompanied by confusion, uncertainty, iterative refinement and delayed understanding. That principle now shapes his product philosophy.
Rather than delivering immediate answers, the company’s AI systems are designed to retain ‘cognitive friction’ and even incorporate ‘dumb’ AI,. The platform asks questions, diagnoses misconceptions, encourages reflection, and guides students through gradual exploration, while allowing for productive failure and struggle.
CocoRobo’s AI-powered teaching platform covers classroom interaction, after-school learning, collaborative knowledge building, and teaching analytics.
Tony has observed a common pattern among the most effective educators: they know when not to use AI.
The teachers who benefit most from the technology are those who set clear boundaries, preserving moments for critical thinking and teacher-student interaction.
Bridging Academic Research and Real Classrooms
In recent years, the company has successfully turned a profit, with its STEM and AI divisions contributing roughly equally to the business. The company now serves more than 400 schools in both Hong Kong and Shenzhen.
Yet the deeper Tony became involved in education, the more convinced he became that keeping pace with technological trends and market demand was not enough.
In 2024, this entrepreneur made an unusual decision: to conduct academic research in the company’s name.
Over the past two years, CocoRobo has collaborated with education researchers from multiple universities, publishing more than a dozen scholarly papers and earning recognition at international academic conferences.
The aim is simple: create a virtuous cycle in which educational research, product design, and classroom outcomes continuously validate one another.


“Academic theory gives us confidence. It also helps us understand where we should be heading.”
As education undergoes rapid transformation, Tony’s primary concern has always been how the next generation will face the future. His answer is remarkably consistent with the values that first inspired him to become an entrepreneur.
Schools should cultivate people who can embrace uncertainty, remain optimistic in the face of challenges, and continue solving problems even when the path forward is unclear.
And if he could offer one piece of advice to aspiring founders, it would be this:
“If you’re hesitating over whether to do something that could genuinely benefit society, my advice is simple—take the chance and do it.”
At a time when the AI EdTech market is still in its infancy, there remains tremendous room for exploration in academic research, business models and technological applications. Drawing on a decade of entrepreneurial experience and a strong network within the education sector, Tony believes that no single individual can drive the industry’s growth alone.
He has observed a growing wave of startups entering the field in recent years, including many one-person companies (OPCs). “These young entrepreneurs are incredibly talented. Rather than competing with them, I would rather collaborate,” he says.
By fostering the open exchange of technology, resources and knowledge, Tony hopes to bring together more like-minded partners to push the boundaries of teaching and learning, and to unlock new possibilities for education in the age of AI.
Edit by: SO Yu Hing and Elok Wong@CUHK Knowledge Transfer Office




