The AI Secretary App is an AI-powered personal secretary application designed to assist users with various tasks through voice and text interactions. It utilizes a local LLM (LLaMA 3 via Ollama) to provide intelligent responses and has memory-based context and retrieval-augmented generation (RAG) capabilities.
- Voice and text input support
- Intelligent responses using LLaMA 3
- Memory management for context-aware interactions
- User-friendly chat interface
- Settings management for user preferences
ai-secretary-app
├── app
│ ├── src
│ │ ├── main
│ │ │ ├── kotlin
│ │ │ │ └── com
│ │ │ │ └── example
│ │ │ │ └── aisecretary
│ │ │ │ ├── MainActivity.kt
│ │ │ │ ├── SecretaryApplication.kt
│ │ │ │ ├── ui
│ │ │ │ │ ├── chat
│ │ │ │ │ │ ├── ChatFragment.kt
│ │ │ │ │ │ ├── ChatViewModel.kt
│ │ │ │ │ │ └── MessageAdapter.kt
│ │ │ │ │ └── settings
│ │ │ │ │ ├── SettingsFragment.kt
│ │ │ │ │ └── SettingsViewModel.kt
│ │ │ │ ├── data
│ │ │ │ │ ├── model
│ │ │ │ │ │ ├── Message.kt
│ │ │ │ │ │ └── ConversationContext.kt
│ │ │ │ │ ├── repository
│ │ │ │ │ │ ├── ChatRepository.kt
│ │ │ │ │ │ └── VoiceRepository.kt
│ │ │ │ │ └── local
│ │ │ │ │ ├── database
│ │ │ │ │ │ ├── AppDatabase.kt
│ │ │ │ │ │ └── dao
│ │ │ │ │ │ └── MessageDao.kt
│ │ │ │ │ └── preferences
│ │ │ │ │ └── UserPreferences.kt
│ │ │ │ ├── ai
│ │ │ │ │ ├── llm
│ │ │ │ │ │ ├── LlamaClient.kt
│ │ │ │ │ │ └── OllamaService.kt
│ │ │ │ │ ├── voice
│ │ │ │ │ │ ├── SpeechRecognizer.kt
│ │ │ │ │ │ └── TextToSpeech.kt
│ │ │ │ │ ├── memory
│ │ │ │ │ │ ├── ConversationMemory.kt
│ │ │ │ │ │ └── MemoryManager.kt
│ │ │ │ │ └── rag
│ │ │ │ │ ├── DocumentStore.kt
│ │ │ │ │ ├── Retriever.kt
│ │ │ │ │ └── VectorStore.kt
│ │ │ │ └── di
│ │ │ │ └── AppModule.kt
│ │ │ ├── res
│ │ │ │ ├── layout
│ │ │ │ │ ├── activity_main.xml
│ │ │ │ │ ├── fragment_chat.xml
│ │ │ │ │ └── fragment_settings.xml
│ │ │ │ ├── values
│ │ │ │ │ ├── colors.xml
│ │ │ │ │ ├── strings.xml
│ │ │ │ │ └── themes.xml
│ │ │ │ └── navigation
│ │ │ │ └── nav_graph.xml
│ │ │ └── AndroidManifest.xml
│ │ └── test
│ │ └── kotlin
│ │ └── com
│ │ └── example
│ │ └── aisecretary
│ │ └── LlmClientTest.kt
│ └── build.gradle.kts
├── build.gradle.kts
├── settings.gradle.kts
└── README.md
- Clone the repository:
git clone https://github.com/A-Akhil/Astra-Ai.git - Open the project in your preferred IDE.
- Copy
secrets.properties.templatetosecrets.propertiesand configure:OLLAMA_BASE_URL=https://your-ollama-server-url LLAMA_MODEL_NAME=llama3:8b
- Ensure you have the necessary SDKs and dependencies installed.
- Build and run the application on an Android device or emulator.
The LLM model is configurable in secrets.properties. For low-RAM systems, use lightweight alternatives:
# Lightweight options (requires less RAM):
LLAMA_MODEL_NAME=phi3
LLAMA_MODEL_NAME=llama3.2:1b
LLAMA_MODEL_NAME=qwen2.5:0.5b
LLAMA_MODEL_NAME=gemma2:2b
LLAMA_MODEL_NAME=llama3:8b- Launch the application and interact with the AI Secretary using voice or text.
- Access settings to customize your preferences.
- The app will remember previous interactions to provide context-aware responses.
Contributions are welcome! Please submit a pull request or open an issue for any enhancements or bug fixes.
This project is licensed under the MIT License. See the LICENSE file for details.
🤖 AI & Machine Learning
-
LLM Integration
- LLaMA 3 integration via Ollama
- System prompt management
- Context-aware responses
- Error handling and retry logic
-
Memory System
- Basic memory storage
- Memory detection from responses
- JSON memory extraction
- Memory cleanup
-
Voice Processing
- Text-to-Speech
- Speech Recognition
- Wake word detection
- Background listening
📱 Core Features
-
User Interface
- Chat interface
- Settings management
- Voice input/output
- Message history
-
System Integration
- Background service
- Lifecycle management
- Model loading/unloading
- Error recovery
🤖 AI & Machine Learning
-
Offline LLM Integration
- On-device model processing
- Model quantization
- Model download management
- Fallback system
-
Enhanced Memory System
- Memory categories
- Memory search
- Memory expiration
- Memory tags
- Export/import feature
-
Learning & Adaptation
- User preference learning
- Response style adaptation
- Conversation history analysis
- Pattern recognition
- Behavior learning
🎤 Voice & Communication
-
Voice Improvements
- Multiple voice options
- Voice activity detection
- Background noise cancellation
- Voice profiles
- Voice command shortcuts
-
Messaging Integration
- WhatsApp integration
- SMS integration
- Telegram integration
- Message scheduling
- Message templates
-
Email Integration
- Gmail/Outlook integration
- Email composition
- Email reading
- Email scheduling
- Email categorization
📱 App Integration & Automation
-
System Integration
- Screen brightness control
- Volume control
- Bluetooth management
- WiFi control
- Battery optimization
-
App Control
- App launching
- Settings management
- Permissions management
- Updates checking
- Usage statistics
-
Quick Actions
- One-tap actions
- Custom shortcuts
- Gesture controls
- Widget controls
- Quick reply templates
📅 Task & Time Management
-
Calendar Integration
- Google Calendar integration
- Meeting scheduling
- Event reminders
- Recurring events
- Calendar sharing
-
Task Management
- Todo list integration
- Task prioritization
- Deadline tracking
- Task sharing
- Progress tracking
🔒 Security & Privacy
-
Access Control
- App-specific permissions
- Data access controls
- Integration permissions
- Privacy settings
- Security policies
-
Data Protection
- End-to-end encryption
- Secure storage
- Data backup
- Data recovery
- Privacy controls
📊 Analytics & Insights
-
Usage Tracking
- App usage statistics
- Integration usage
- Command frequency
- Response times
- Error rates
-
Performance Monitoring
- Battery usage
- Memory usage
- CPU usage
- Network usage
- Storage usage
🔄 Integration & APIs
-
Third-party Apps
- Slack integration
- Microsoft Teams
- Zoom integration
- Trello integration
- Jira integration
-
Cloud Services
- Google Drive
- Dropbox
- OneDrive
- iCloud
- Backup services
🎨 UI/UX Improvements
-
Customization
- Dark/light theme
- Custom voice commands
- Custom shortcuts
- Custom templates
- Custom workflows
-
Accessibility
- Voice control
- Gesture control
- Screen reader support
- High contrast mode
- Font size adjustment
Current Development
-
Voice Improvements
- Basic TTS implementation
- Basic Speech Recognition
- Multiple voice options
- Voice activity detection
- Background noise cancellation
-
Memory System
- Basic memory storage
- Memory detection
- Memory categories
- Memory search
- Memory expiration
-
UI/UX
- Basic chat interface
- Settings screen
- Dark/light theme
- Custom voice commands
- Gesture controls