Swasthya Initiative - AI-Driven Student Mental Health Assessment

Abstract

Multi-Agent Artificial Intelligence systems are increasingly applied to complex, domain-specific problems requiring continuous monitoring, contextual understanding, and coordinated intervention. This paper presents Swasthya Initiative, a multi-agent AI framework for student mental health assessment and support built using React, Node.js, Express, MySQL, and AI APIs such as Claude and HuggingFace. The system comprises multiple specialized modules — an Assessment Agent that evaluates stress, anxiety, and depression scores from structured questionnaires, a Risk Detection Agent that classifies severity levels and triggers alerts based on defined thresholds, a Support Recommendation Agent that assists in counselor assignment and follow-up tracking, and an AI Chatbot Agent that provides empathetic, context-aware responses using layered NLP and LLM-based reasoning. A rule-based Journal Analysis module classifies emotional tone using keyword-based sentiment scoring. Secure user authentication is implemented using JWT tokens and bcrypt hashing, ensuring privacy and role-based access for students and teachers. A central logic controller manages workflow transitions across modules and maintains the intervention lifecycle. Experimental results demonstrate effective early detection, improved intervention tracking, and meaningful user engagement, validating the practical applicability of AI-driven systems in student mental health monitoring and support.

Introduction

The Swasthya Initiative is an AI-driven student mental health monitoring and support system designed to address the growing prevalence of stress, anxiety, and depression among students. Traditional counseling services and existing digital mental health platforms often lack continuous monitoring, personalization, and proactive intervention. To overcome these limitations, the proposed system combines structured mental health assessments, rule-based risk analysis, Retrieval-Augmented Generation (RAG), and Large Language Models (LLMs) to provide personalized, timely, and context-aware support.

The system is built on a three-tier architecture consisting of a React frontend, Node.js/Express backend, and MySQL database. It employs multiple specialized modules, including assessment, risk detection, support management, and logging modules, to evaluate user responses, calculate standardized stress, anxiety, and depression scores, identify high-risk cases, and facilitate intervention through counselor assignment and AI-assisted support. Semantic retrieval techniques using sentence transformers enhance the accuracy and relevance of chatbot responses.

The system operates through three primary workflows: Assessment Flow, Intervention Flow, and Support Flow, enabling continuous mental health monitoring and personalized assistance. Scenario-based testing on diverse student profiles demonstrated effective risk detection, successful intervention tracking, and real-time responsiveness, with average assessment and support response times of 1.75 seconds and 3.85 seconds, respectively.

Results indicate that the modular architecture improves scalability, reliability, and adaptability compared to traditional survey-based and chatbot-only approaches. While the system effectively identifies high-risk mental health conditions and supports ongoing monitoring, its accuracy depends on self-reported user data and it cannot replace professional clinical diagnosis. Future enhancements include integrating advanced clinical support mechanisms and more sophisticated decision-making models to improve intervention quality and reliability.

Conclusion

This research has presented Swasthya Initiative, a modular AI-driven framework designed to bridge the gap between traditional mental health support systems and the dynamic, evolving needs of student well-being. By integrating structured analysis with intelligent AI-driven insights, we have shown that it is possible to provide context-aware, reliable mental health insights without constant dependence on human intervention. The core strength of our approach lies in the separation of responsibilities—assigning dedicated modules to handle tasks such as assessment processing, risk detection, and support management—which significantly improves scalability and reduces inconsistencies in decision-making. Our evaluation indicates that the Swasthya Initiative offers a more adaptive and responsive experience compared to traditional survey-based systems or simple rule-based platforms. By maintaining a continuous record of user assessments and behavioral patterns, the system progressively refines its insights, aligning them with individual mental health trends and support needs. This transition from reactive support to proactive, data-driven intervention represents a meaningful advancement in student mental health management. However, moving from a research prototype to a fully deployable system requires further enhancement. The system remains dependent on the accuracy of user-provided data, and the absence of direct clinical validation mechanisms is an important limitation. Future improvements will focus on integrating validated mental health frameworks and strengthening connections with institutional support systems. Ultimately, the Swasthya Initiative serves as a strong proof-of-concept for scalable, intelligent, and context-aware mental health monitoring and support in academic environments.

References

[1] Fitzpatrick et al., “Delivering Cognitive Behavioral Therapy via AI Chatbot (Woebot),” JMIR Mental Health, 2017. [2] Inkster et al., “Wysa: AI-Based Mental Health Support System,” JMIR mHealth, 2018. [3] Torous et al., “Digital Mental Health and Mobile Applications,” NPJ Digital Medicine, 2020. [4] Miner et al., “Conversational Agents in Healthcare: Safety and Effectiveness,” NPJ Digital Medicine, 2016. [5] World Health Organization (WHO), “Mental Health and COVID-19,” 2022. [6] National Institute of Mental Health (NIMH), “Mental Health Statistics,” 2023. [7] American Psychological Association (APA), “Student Mental Health Trends,” 2023. [8] Beck et al., “Depression Assessment and Clinical Scales,” 1996. [9] Spitzer et al., “Generalized Anxiety Disorder Scale (GAD-7),” 2006. [10] FastAPI Framework, 2024.

Copyright

Copyright © 2026 Byagari Sahasra Lahari, C Nymisha Reddy, Guntuka Rishika, Sriram Abhigna. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.