MYRA AI - Voice-Controlled Android Assistant 🎙️

MYRA AI is an advanced voice-controlled Android assistant that can control your mobile device, handle phone calls, edit media, and answer complex questions using state-of-the-art AI models.

Features

🎤 Voice Control

• Speech-to-Text: Real-time voice recognition with partial results
• Text-to-Speech: Natural voice responses
• Wake Word Detection: Say "Hey MYRA" to activate hands-free
• Continuous Listening Mode: Always-on voice assistant

🧠 Dual AI Engine

• Gemini API (Fast): Quick responses for simple tasks, device control, and voice commands
• OpenRouter API (Powerful): Complex reasoning using free top models:
  • OpenAI GPT-OSS-120B (deep reasoning)
  • DeepSeek V4 Flash (code & analysis)
  • Qwen3-72B (creative writing)
  • DeepSeek R1 (fallback)

📱 Device Control

• Open any app by voice command
• Make phone calls
• Set alarms and timers
• Toggle WiFi, Bluetooth, and other settings
• Adjust volume and brightness
• Open URLs and search the web
• Play music via Spotify or YouTube Music

📞 Call Handling

• Answer calls with AI: MYRA can pick up calls and speak to the caller
• Speak during calls: Send voice messages through active calls
• Incoming call detection: Get notified and let AI handle calls
• Call management: Hold, unhold, and end calls via voice

💬 Quick Messaging

• Send SMS to any contact by voice
• Contact name resolution (say the name, not the number)
• Support for long messages (auto-split)
• Read recent messages

🖼️ Image Processing

• Format Conversion: PNG ↔ JPEG ↔ WebP
• Resize: Scale images to any dimension
• Rotate: Rotate by any degree
• Crop: Crop to specific regions
• Filters: Grayscale, Sepia, Invert, Brightness, Contrast, Warm, Cool
• Compress: Reduce file size with quality control

🎬 Video Editing

• Trim: Cut videos to specific time ranges
• Merge: Combine multiple videos into one
• Rotate: Rotate video by 90°, 180°, 270°
• Speed: Change playback speed (slow-mo or fast-forward)
• Contrast: Adjust video contrast
• Remove Audio: Mute video
• Extract Audio: Get audio track from video

Architecture

app/src/main/java/com/myra/ai/
├── MainActivity.kt              # Main UI with Jetpack Compose
├── App.kt                       # Application class
├── ai/
│   ├── AIModelRouter.kt         # Intelligent routing between Gemini & OpenRouter
│   ├── GeminiService.kt         # Gemini API for quick responses
│   ├── OpenRouterService.kt     # OpenRouter API for complex tasks
│   ├── DeviceController.kt      # Device control execution
│   ├── ToolExecutionManager.kt  # Orchestrates all capabilities
│   ├── ConversationEngine.kt    # Conversation management
│   ├── IntentEngine.kt          # Intent parsing
│   └── ...
├── voice/
│   ├── VoiceAssistantService.kt # Foreground voice service
│   └── VoiceActivationManager.kt # Wake word & continuous listening
├── call/
│   ├── CallHandlerService.kt    # InCallService for speaking in calls
│   └── IncomingCallReceiver.kt  # Incoming call detection
├── communication/
│   └── SMSManager.kt            # SMS sending & contact resolution
├── media/
│   ├── ImageProcessor.kt        # Image conversion & editing
│   └── VideoEditor.kt           # Video editing with Media3
├── agents/                      # AI agent modules
├── emotion/                     # Emotional AI system
├── memory/                      # Long-term memory
├── overlay/                     # Floating orb overlay
├── security/                    # Security vault
└── vision/                      # Vision & OCR

Setup Instructions

Prerequisites

• Android Studio Hedgehog (2023.1.1) or later
• JDK 17
• Android SDK 34
• Kotlin 2.0.0

API Keys Setup

1. Gemini API Key: Get from Google AI Studio
2. OpenRouter API Key: Get from OpenRouter (free tier available)

Configuration

1. Clone the repository:

git clone https://github.com/anxcrn/Myra-ai-assistant.git
cd Myra-ai-assistant

2. Open in Android Studio

3. Add your API keys in MainActivity.kt:

const val GEMINI_API_KEY = "your_gemini_api_key_here"
const val OPENROUTER_API_KEY = "your_openrouter_api_key_here"

Tip: For production, store API keys in local.properties or use BuildConfig fields.

4. Build and run on your Android device (API 26+)

Permissions

The app requires the following permissions (requested at runtime):

• Microphone: For voice recognition
• Phone: For making and handling calls
• SMS: For sending text messages
• Contacts: For contact name resolution
• Storage/Media: For image and video processing
• Camera: For taking photos
• System Alert Window: For floating overlay

Usage Examples

Voice Commands

• "Open YouTube"
• "Call Mom"
• "Text John I'll be there in 10 minutes"
• "Set alarm for 7 AM"
• "Turn on WiFi"
• "Play some jazz music"
• "Convert this image to PNG"
• "Trim the video from 10 seconds to 30 seconds"

Complex Queries (uses OpenRouter)

• "Explain quantum computing in simple terms"
• "Write a Python function to sort a linked list"
• "Compare React and Vue.js for a new project"
• "Write a short story about a robot learning to paint"

Tech Stack

Component	Technology
Language	Kotlin 2.0
UI	Jetpack Compose + Material 3
AI (Fast)	Google Gemini 2.0 Flash
AI (Complex)	OpenRouter (GPT-OSS-120B, DeepSeek V4)
DI	Hilt
Networking	Retrofit + OkHttp
Media	Media3 Transformer
Database	Room
Async	Coroutines + Flow

Roadmap

• Multi-language support (Telugu, Hindi, etc.)
• Custom wake word training
• Screen automation via Accessibility Service
• WhatsApp/Telegram integration
• Smart home device control
• Offline mode with on-device models
• Widget for quick voice access

License

This project is for personal use. All rights reserved.

---

Built with ❤️ by MYRA AI Team