Skip to content

StreAM: An LLM-Based System for Stress-Adaptive Meditation at Work

Luca L. Bennardo, Luisa Weber, Jonas Hummel, Martin Feick

CHI Extended Abstracts April 13, 2026 DOI: 10.1145/3772363.3798694 (opens in new tab)

Study at a glance

AI-extracted from the abstract
Characteristics Theoretical or philosophical paper Peer reviewed
Intervention StreAM (Stress-Adaptive Meditation)
Topics Meditation
Key points Proposes StreAM, a system combining real-time heart rate variability monitoring with large-language-model-guided micro-meditation that adapts to user preferences and continues until physiological signs of stress reduction appear, as a scalable, unobtrusive approach to workplace stress management.

Abstract

Workplace stress is a critical factor that undermines employee performance and leads to physical ailments and mental health challenges. Although meditation is widely recognized as an effective stress-reduction strategy, traditional meditation practices are often too time-consuming or complicated to be easily incorporated into professional settings. To address this issue, we introduce StreAM (Stress-Adaptive Meditation), a system that combines real-time stress detection with personalized micro-meditation exercises. StreAM continuously monitors heart rate variability to identify heightened stress levels and initiate an adaptive meditation session guided by a large language model. These sessions are tailored to users’ personal preferences and persist until physiological indicators signal stress reduction, enabling individuals to swiftly regain focus. By leveraging the growing potential of heart rate sensors in everyday devices, StreAM demonstrates a pathway toward scalable, accessible, and unobtrusive workplace stress management, potentially enhancing employee well-being and fostering healthier, more productive work environments.

Comparable studies

Other theoretical and historical papers on meditation, most cited first.

Citations in the library

Explore topics