The AI Wellness Quest hackathon from Sberbank gathered 17 teams who had to study the "Lose Weight for Health" application in 8 hours and propose new features with a working prototype. A team of first—year students from the National Research Nuclear University MEPhI Avalon (Daniil Babkin, Alexander Pirozhok, Peter Zubarev, Maxim Lobanov, Ilya Podmarkov) took the first place, suggesting embedding an emotional layer in the application - mood tracking, which the AI assistant connects with user behavior.

At the hackathon, our students showed by the example of nutrition that the mood heat map reveals the triggers of overeating. But the approach itself is universal — emotions can be associated with sleep quality, activity, water balance, and any habit. When the AI captures a difficult period, it offers personal support — a movie, music, podcast tailored to the user's interests. And the streaking system, by analogy with Duolingo, turns daily useful actions into a game and gives a reason to return to the application every day. The solution turned out to be unique, which the jury liked. As a result of the hackathon, the app developers expanded the product concept: from a weight loss tool, it transforms into an AI companion that helps form healthy habits and becomes an everyday assistant in improving the quality of life.