MindEase Platform: An AI-Based, Privacy-Preserving Mental Health Support System with Emotion-Aware Chat, Mood Tracking and Crisis Detection

Access to timely mental health support remains limited, and individuals may hesitate to share personal concerns with others or cloud-based services. This paper presents MindEase, a full-stack mental wellness platform that integrates an AI-powered conversational assistant, daily mood check-ins, journaling with emotion analysis, guided relaxation activities, and a safety-focused crisis-detection mechanism. The platform combines a React.js frontend, a Node.js/Express REST API, a PostgreSQL database, and a modular natural language processing (NLP) engine. The chatbot uses keyword-based emotion detection, tokenisation, stemming, and sentiment analysis to generate supportive responses without relying on external conversational AI APIs. For journal and voice analysis, the system incorporates pretrained models for text emotion detection, speech transcription, and audio emotion recognition. A dedicated risk-screening mechanism identifies potentially concerning language and prioritises safety-oriented responses, including relevant crisis-support information. The platform also provides mood tracking, emotional insights, and activities to support self-awareness and relaxation. This paper presents the system architecture, technologies, NLP methodology, database design, API structure, development workflow, and testing approach for evaluating functionality, chatbot behaviour, and security.