Well-being Companion Chatbot

Table of Contents

• About This Project
• Features
• Installation and Setup
  • Prerequisites
  • Environment Variables
  • Running the Application
• Usage
• What Makes it Wonderful?
• Future Enhancements
• Troubleshooting

About This Project

🎥 Demo Video

✨ Click the image to watch the calming AI chatbot in action!

The Well-being Companion Chatbot is an AI-powered conversational agent designed to provide emotional support and engage in meaningful interactions focused on mental and emotional well-being. Built with Flask for the web interface and leveraging Google's Gemini models, it offers a personalized and private chat experience.

Features

• AI-Powered Conversations: Utilizes advanced large language models (Google Gemini 1.5 Flash) to provide empathetic and supportive responses.
• User Privacy: Designed with user privacy in mind, ensuring that individual conversation histories are kept separate and private.
• Audio Input (Speech-to-Text): Allows users to interact with the chatbot using their voice, converting speech to text for seamless communication (requires HTTPS for mobile browsers).
• Conversation Management: Users can start new chats, load previous conversations, view a list of their saved conversations, and delete specific conversation histories.
• Guided Breathing Exercises: Offers a structured, guided breathing exercise to help users relax and manage stress.
• Mood Analysis (Planned): Future capabilities may include analyzing user sentiment to tailor responses more effectively.
• Responsive Web Interface: A user-friendly web interface built with HTML, CSS, and JavaScript, designed for a smooth experience across various devices.
• Conversation Persistence: User conversation data is stored to maintain continuity across sessions. Note: For production deployment, a robust, persistent database solution is recommended for data integrity and scalability.

Installation and Setup

Follow these steps to get your Well-being Companion Chatbot up and running locally.

Prerequisites

• Python 3.8+
• pip (Python package installer)
• FFmpeg: Essential for audio transcription.
  • Windows: Download a pre-built binary from ffmpeg.org/download.html and extract it. Update the ffmpeg_bin_path in web_app.py to point to the bin directory within your FFmpeg installation.
  • macOS: brew install ffmpeg
  • Linux: sudo apt update && sudo apt install ffmpeg

Environment Variables

Create a .env file in the root directory of your project and add the following:

GEMINI_API_KEY="YOUR_GOOGLE_GEMINI_API_KEY"
TWINWORD_API_KEY="YOUR_TWINWORD_RAPIDAPI_KEY"

• GEMINI_API_KEY: Obtain this from the Google AI Studio.
• FLASK_SECRET_KEY: Generate a strong, random string for Flask session security (e.g., using secrets.token_urlsafe(32) in Python). For deployment on platforms like Render, ensure this is set as an environment variable in your service settings.
• TWINWORD_API_KEY: Obtain this from RapidAPI. You will need to sign up for an account and subscribe to the Twinword Sentiment Analysis API to get your key.

Running the Application

1. Clone the repository:

   git clone https://github.com/DohaSK/ai-chatbot-well_being.git
   cd ai-chatbot-well_being

   (Note: If you are already in the project directory, skip git clone and cd.)

2. Install dependencies:

   pip install -r requirements.txt

   (A requirements.txt file is needed for this. If it's missing, you can create it with pip freeze > requirements.txt after installing dependencies.) The necessary packages are likely: flask, whisper-openai, python-dotenv, google-generativeai, textblob, requests.

3. Start the Flask application:

   python web_app.py

Usage

• Chat: Type your messages in the input box and press Enter or click the send button.
• Voice Input: Click the microphone icon to start/stop recording your voice. Your speech will be transcribed and sent to the chatbot.
• Breathing Exercise: Click the "Start Breathing Exercise" button to receive guided steps for relaxation.
• Conversation History: Use the sidebar menu (toggle button in the header) to view, load, or delete previous conversations.
• New Chat: Click the "New Chat" button in the sidebar to start a fresh conversation.

What Makes it Wonderful?

This chatbot stands out as a wonderful companion due to its focus on well-being and privacy. Unlike many general-purpose chatbots, it's specifically tailored to offer empathetic and supportive interactions. The key wonderful aspects include:

• Genuine Empathy (AI-driven): Powered by Gemini, it strives to understand and respond to user emotions with care and appropriate support, making conversations feel more natural and helpful for mental wellness.
• Unwavering Privacy: The chatbot respects user privacy, providing a confidential and secure environment for personal interactions.
• Accessibility through Voice: Voice input makes it highly accessible and convenient, allowing users to interact naturally without typing, which is particularly beneficial for those who prefer speaking their thoughts.
• Ease of Use: The simple and intuitive web interface, combined with features like conversation management and guided exercises, makes it easy for anyone to engage with and benefit from the companion.
• Continuous Improvement: As an AI, it's designed to continuously improve based on interactions (though this project focuses on foundational interaction rather than direct model retraining from user input).

It's designed to be a supportive digital friend, available whenever needed, offering a judgment-free space to explore thoughts and feelings.

Future Enhancements

• Mood Analysis Integration: Enhance the chatbot with more sophisticated mood detection and tailored conversational flows based on user sentiment.
• User Authentication: Implement user accounts and login for more personalized experiences and potentially richer profile management.
• Broader Well-being Tools: Add more guided exercises, journaling features, or connections to external well-being resources.
• Speech Synthesis (Text-to-Speech): Allow the chatbot to speak its responses back to the user for an even more immersive voice experience.
• Frontend Refinements: Enhance the UI/UX for a more polished and engaging user experience.

Troubleshooting

• "Error accessing microphone: NotAllowedError" or "PermissionDeniedError" on mobile (especially iOS): Ensure your deployed website is accessed via HTTPS. Microphone access is a sensitive feature usually restricted to secure contexts.
• "ModuleNotFoundError: No module named...": Ensure all dependencies are installed (pip install -r requirements.txt).
• "404 models/gemini-pro is not found...": Update the chat_model_name and title_model_name in web_app.py to use a model name available to your API key (e.g., gemini-1.5-flash). Refer to the console output when the app starts for a list of available models.
• FFmpeg related errors: Verify that FFmpeg is installed and that the ffmpeg_bin_path in web_app.py is correctly set to your FFmpeg bin directory.
• "I encountered an error: module 'google.api_core.exceptions' has no attribute 'BlockedPromptException'": Ensure BlockedPromptException is imported directly from google.generativeai.types in Core/chatbot.py.
• Git Issues (merge conflicts, unable to push/pull): Follow standard Git conflict resolution. For stubborn issues, consider stashing local changes, performing a git reset --hard origin/main, reapplying the stash, and then pushing. Ensure .gitignore correctly lists debug.log, Core/conversation_memory.json, Core/main_backup.py, and flask_session/.