🎤 Customizable Voice AI Assistant (Misa AI)

This Python project is a fully voice-controlled AI assistant, powered by:

• Google Gemini 1.5 Flash natural language generation
• ElevenLabs for high-quality, customizable text-to-speech
• SpeechRecognition for capturing your voice input
• Pygame for audio playback

---

✅ Features

• 🎙️ Voice-based input using microphone
• 🧠 AI-generated responses with customizable tone/personality
• 🔊 Realistic speech synthesis with ElevenLabs
• 🔁 Continuous conversation loop
• 🔧 Highly customizable via .env file

---

🛠️ Setup Instructions

1. Install Dependencies

pip install -r requirements.txt

Required packages:

SpeechRecognition
pygame
python-dotenv
google-generativeai
elevenlabs

2. Configure Environment Variables

Create a .env file in your project directory with the following:

GEMINI_API_KEY=your_google_gemini_api_key
ELEVENLABS_API_KEY=your_elevenlabs_api_key

AI_NAME=Misa
AI_PERSONALITY=You are a helpful AI assitant.
VOICE_ID=21m00Tcm4TlvDq8ikWAM
MODEL_ID=eleven_multilingual_v2
EXIT_COMMANDS=exit,quit,stop,goodbye,bye,exit misa
MAX_RETRIES=5

You can find available voice_ids by logging into ElevenLabs and checking your voice dashboard.

---

🚀 Usage

Run the Assistant

python main.py

Conversation Flow

1. The assistant will listen to your voice input via microphone.
2. The input is converted to text using Google Speech Recognition.
3. Gemini generates a custom AI response based on your conversation history and personality settings.
4. ElevenLabs converts the AI text into spoken audio and plays it.
5. The loop continues until you say any of the exit commands (e.g., "exit", "stop").

---

🔧 Customization Options

All customization is controlled via environment variables:

Variable	Description
AI_NAME	Name of your assistant (used in printed and spoken text)
AI_PERSONALITY	Prompt that defines your assistant’s personality and tone
VOICE_ID	ElevenLabs voice ID (e.g., Rachel, Bella, etc.)
MODEL_ID	ElevenLabs voice model ID (default: eleven_multilingual_v2)
EXIT_COMMANDS	Comma-separated phrases to stop the assistant
MAX_RETRIES	Number of retries for failed voice recognition attempts

---

🧠 How It Works

Your voice → [SpeechRecognition] → Text
          → [Google Gemini] → AI response (text)
          → [ElevenLabs] → Audio (MP3)
          → [Pygame] → Plays the voice

Conversation context is preserved to maintain continuity and sarcasm 😎.

---

🧪 Tips & Troubleshooting

• 🔇 Microphone not working? Make sure your default input device is enabled and accessible.
• 🔑 API Key errors? Double-check your .env keys and ensure the services are active.
• 🧠 Want a polite assistant? Just replace the AI_PERSONALITY with a friendly or formal description.

---

📁 Project Structure

├── main.py
├── .env
├── requirements.txt

---

📄 License

MIT License — feel free to use, fork, and personalize.

---