Zhu Shuhan, a 13-year-old from Shanghai, has developed an AI-based app called Zai Chang ("On the Spot") that helps children focus on their homework. The project, presented at the Markethon competition in Hangzhou, demonstrated impressive market results: 86 orders were placed in just three days, generating revenue of ¥18,295 (approximately $2,700).
The service works using computer vision. The device's camera monitors the child's behavior: if they get distracted, leave their workspace, or pick up their phone, a "virtual parent" appears on screen with a reminder to return to their studies. At the end of the session, the app generates a detailed report on concentration levels.
Of particular interest is the mechanics of creating the digital character. Simply upload a photo of a parent, and the algorithm generates a virtual copy of them. Zhu emphasizes that the images are deleted immediately after analysis, and no continuous video recording takes place — an important aspect from a privacy standpoint.
Personal Experience as a Starting Point
The idea came from her own experience: Zhu's parents work a lot, and she often has to do her homework on her own. She admits that in such moments, she herself often got distracted by her phone. A survey of classmates confirmed that the problem is widespread. The first tester was her younger brother, who was preparing to enter elementary school.
Notably, the initial version included a "angry mom" feature and sending notifications to parents about their child's inattention. However, the young developer abandoned these options, deciding that the app should be gentler and leave the child personal space. "I don't like tattling," she explained.
Technology Without Formal Education
Zhu began studying AI about a year ago, independently learning through educational platforms and online tools. She has virtually no traditional programming skills: all the code was generated by a large language model based on natural language queries. The girl described tasks, tested features, and fixed errors, asking the chatbot about anything she didn't understand.
At the competition, where 59 teams competed in real sales, the project took third place. However, the working version was only ready a day before the event, and technical hiccups arose during the demonstration — from forgotten account credentials to delays in system alerts. Despite this, the commercial success was significant: a yearly subscription cost ¥99 (~$15), and a lifetime one ¥199 (~$30).
More than 60 orders came in online, with the rest through personal sales and live streams on social media, where the girl's mother helped with promotion. After the competition, Zhu continues to refine the service, analyzing statistics through the admin panel and gathering feedback on WeChat.
This case is telling for the industry: it demonstrates how the accessibility of generative AI democratizes product development. In China, where a national plan to integrate AI into education was recently launched, such initiatives could become a trend. However, it's worth noting that niche solutions for attention monitoring face questions of ethics and trust — and Zhu's approach, which rejected "tattling" in favor of gentle reminders, seems more mature than that of many adult developers.