MindfulAgents: Personalizing Mindfulness Meditation via an Expert-Aligned Multi-Agent System
This Paper in Plain English
MindfulAgents explores how AI can make meditation apps feel more personal and useful. Instead of giving everyone the same guided meditation, the system uses a team of AI agents to understand a person's current needs, encourage reflection, and create meditation guidance that better fits the moment.
You Today
Mood, stress, time, and what kind of support would help.
A Session That Fits
Guided meditation shaped around the moment, not a generic script.
In two studies, people who used MindfulAgents were more engaged with meditation, felt more self-aware, and reported less stress during sessions. Over four weeks, participants also kept using meditation more consistently and showed gains in mindfulness. The main takeaway is that carefully designed AI personalization may help people build a meditation habit that feels relevant to their real lives.
Discussion Takeaways
The interesting shift is not just smarter recommendations. The system can start to feel like a supportive companion when it remembers context and responds with care.
The reflection step acts like a bridge between daily life and formal practice, helping people connect what they feel before a session to what they notice afterward.
Expert-approved structure keeps the meditation grounded, but too much repetition can make people disengage. The paper points toward safe building blocks that can be recombined in new ways.
A strong long-term version would not just adapt one session. It would help users see their practice becoming part of who they are.
Abstract
Mindfulness meditation is a widely accessible and evidence-based method for supporting mental health. Despite the proliferation of mindfulness meditation apps, sustaining user engagement remains a persistent challenge. Personalizing the meditation experience is a promising strategy to improve engagement, but it often requires costly and unscalable manual effort.
We present MindfulAgents, a multi-agent system powered by large language models that (1) generates guided meditation scripts based on an expert-established mindfulness framework, (2) encourages users' reflection on emotional states and mindfulness skills, and (3) enables real-time personalization of the mindfulness meditation experience for each user.
In a formative lab study (N=13), MindfulAgents significantly improved in-session engagement (p = 0.011) and self-awareness (p = 0.014), and reduced momentary stress (p = 0.020). Furthermore, a four-week deployment study (N=62) demonstrated a notable increase in long-term engagement (p = 0.002) and level of mindfulness (p = 0.023). Participants reported that MindfulAgents offered more relevant meditation sessions personalized to individual needs in various contexts, supporting sustained practice.
Our findings highlight the potential of LLM-driven personalization for enhancing user engagement in digital mindfulness meditation interventions.
Key Figure
Figure 1: MindfulAgents system overview
BibTeX
@inproceedings{wu2026mindfulagents,
title={MindfulAgents: Personalizing Mindfulness Meditation via an Expert-Aligned Multi-Agent System},
author={Wu, Mengyuan and Jiang, Zhihan and Fan, Yuang and Feng, Richard and Dharmavaram, Sahiti and Polowitz, Mathew and Fallon, Shawn and Islam, Bashima and Benson, Lizbeth and Tung, Irene and Creswell, David and Xu, Xuhai},
booktitle={Proceedings of the 2026 CHI Conference on Human Factors in Computing Systems},
series={CHI '26},
pages={1--25},
year={2026},
publisher={ACM},
address={New York, NY, USA},
location={Barcelona, Spain},
doi={10.1145/3772318.3791817},
url={https://doi.org/10.1145/3772318.3791817}
}Coming Soon in Equa App
Sign up to become a beta user and be among the first to try the latest Equa experience, improved directly from our CHI research findings.