Authors - Abhishek Chaudhari, Mahalakshmi Bodireddy, Aditya Bhor, Onkar Dadas, Prajakta Shinkar, Chinmay Chougule Abstract - The growing mental health challenges around the globe need access to scalable, available, and safety conscious digital interventions. The paper describes a mental health support platform, based on AI, which combines conversational intelligence, multi-therapeutic persona modeling, structured mood analytics, proactive crisis identification, multi-lingual interaction, and voice-based access in a secure full stack design. The system, which runs on the Google Gemini AI, provides context-sensitive therapeutic dialogue and performs four-dimensional mood analysis of anxiety, stress, depression, and wellbeing, allowing longitudinal assessment by providing interactive dashboards and automated reporting. A safety-first crisis override system offers validated emergency capacity in the high-risk situations. The platform also includes multilingual voice feedback to facilitate inclusion of the visually impaired users and non-English speaking communities in providing inclusive digital mental health care. The proposed system is capable of changing the prevalent perception that AI and its applications may never be responsible and scalable because it integrates therapeutic diversity, structured analytics, accessibility features, and proactive safety controls into a single framework.