• Project Overview:
Wellness AI Chatbot: I developed an innovative AI-powered wellness chatbot designed to offer personalized nutrition tips and relaxation guidance. The primary objective was to create an engaging and supportive virtual assistant that interacts naturally and provides valuable wellness information.
• Key Technologies and Features:
- Natural Speech Synthesis and Real-time Speech Recognition:
StyleTTS2: Implemented for natural and expressive speech synthesis, enabling the chatbot to communicate with users in a human-like and engaging manner.
FasterWhisper: Utilized for real-time speech recognition, allowing the chatbot to understand and respond to voice inputs efficiently.
- Voice Interaction and Audio Feedback:
Voice to Text Conversion: Users can record their voice, which is then converted to text, enabling hands-free interaction.
Text to Audio Responses: The chatbot converts its text responses back into audio files, creating a seamless and interactive user experience.
- Information Retrieval and Utilization:
RAG (Retrieval-Augmented Generation): Used to retrieve relevant information from a comprehensive database of nutrition and hypnotic scripts, ensuring responses are accurate and contextually relevant.
Mistral LLM: Integrated to enhance the chatbot's ability to provide detailed and accurate responses by leveraging large language models trained on specific wellness-related content.
• Development and Deployment:
Google Colab: Developed the project on Google Colab, taking advantage of its powerful computational resources.
Streamlit: Hosted the chatbot on a Streamlit web app, providing an accessible and interactive platform for users to engage with the chatbot.
• Impact and Benefits:
Enhanced User Engagement: Advanced speech synthesis and recognition technologies offer a more interactive and user-friendly experience.
Personalized Wellness Guidance: Provides tailored nutrition tips and relaxation techniques, catering to individual user needs and preferences.
Accurate and Contextual Responses: Ensures users receive relevant and reliable information, enhancing the chatbot's effectiveness as a wellness support tool.
GPU Constraints: One of the major obstacles was the lack of sufficient GPU resources, which impacted the chatbot's performance and response times. This limitation highlighted the need for more robust computational infrastructure for real-time applications.