Soul Support : AI Driven Emotional Assistance ChatBot

This paper introduces the development of an AI-based Mental Health Therapist Chatbot designed to provide immediate, accessible emotional support and mental health information. Utilizing deep learning and natural language processing (NLP), the chatbot identifies user intents and delivers appropriate responses across various mental health topics, including anxiety, stress, and sadness. The system is deployed via a web interface that supports both text and voice input, enhancing user accessibility. On the backend, the chatbot uses a Flask server, a trained Keras model for intent classification, and preprocessing pipelines built with NLTK and SpaCy. Experimental evaluations show high accuracy in intent recognition and efficient real-time interaction. Additionally, user feedback reflects a positive reception in terms of helpfulness and ease of use. While the chatbot does not replace professional care, it acts as a supportive digital tool for initial engagement and emotional guidance. Future enhancements will focus on multilingual support, contextual dialogue, and integration with mental health services.