Visulo is a voice-first AI tutor that sees what you're studying. Capture screenshots and voice notes, then ask questions about your content using real AI models.

• 📸 Screenshot Capture: Capture any screen content with OCR text extraction
• 🎤 Voice Notes: Push-to-talk voice recording with transcription
• 🧠 AI-Powered Answers: Get grounded answers from your captured content
• 🔍 Semantic Search: Find relevant information using vector embeddings
• ⚙️ Flexible Providers: Choose between local and cloud AI providers

• Local Python (Default): Uses sentence-transformers locally (privacy-focused)
• OpenAI: Cloud-based embeddings (requires API key)
• Stub: Development/testing mode

• Ollama (Default): Local LLM via Ollama (privacy-focused)
• OpenAI GPT-3.5: Cloud-based LLM (requires API key)
• Stub: Development/testing mode

• Node.js (v16 or higher)
• Rust (latest stable)
• Python 3.7+ (for local embeddings)

1. Clone the repository

   git clone <repository-url>
   cd visulo

2. Install dependencies

   npm install

3. Set up local embeddings (optional but recommended)

   python setup_local_embeddings.py

4. Set up Ollama (optional but recommended)

   # Install Ollama from https://ollama.ai
   # Then pull a model:
   ollama pull llama3

5. Run the development server

   npm run tauri:dev

1. Local Embeddings:

   • Run python setup_local_embeddings.py to install sentence-transformers
   • In Settings → Embedding Provider → Select "Local Python"

2. Local LLM:

   • Install Ollama
   • Pull a model: ollama pull llama3
   • In Settings → LLM Provider → Select "Ollama"

1. Get an OpenAI API key from OpenAI Platform
2. In Settings → Enter your API key
3. Select "OpenAI" for both Embedding and LLM providers

1. Capture Content:

   • Click the camera button or press F9 for screenshots (works globally)
   • Hold the microphone button or press Ctrl+Space for voice notes

2. AI Processing:

   • OCR extracts text from screenshots
   • Content is automatically indexed with embeddings
   • AI generates contextual answers

3. Ask Questions:

   • Your captured content becomes searchable
   • AI provides grounded answers with citations
   • View source snippets for each answer

• F9: Quick screenshot (works globally, even when app is minimized)
• Ctrl+Space: Push-to-talk (hold to record)

• Capture System: Screenshot capture, OCR processing
• Indexing Service: Text chunking, embedding generation, SQLite storage
• Retrieval Service: Semantic search, answer generation
• Provider System: Modular AI providers (local/cloud)

• Capture Interface: Screenshot and voice capture UI
• History Sidebar: Real-time processing indicators
• Settings Panel: Provider configuration
• Toast System: User feedback and notifications

Screenshot/Voice → OCR/Transcription → Text Chunks → Embeddings → SQLite
                                                                      ↓
Query → Query Embedding → Similarity Search → Top-K Chunks → LLM → Grounded Answer

visulo/
├── src/                    # React frontend
│   ├── components/         # UI components
│   └── styles/            # Tailwind CSS
├── src-tauri/             # Rust backend
│   └── src/               # Tauri application logic
├── public/                # Static assets
└── setup_local_embeddings.py  # Setup script

# Development
npm run tauri:dev

# Production build
npm run tauri:build

1. Embedding Provider:

   • Add variant to EmbeddingProvider enum
   • Implement in EmbeddingService::generate_embedding()
   • Update frontend settings UI

2. LLM Provider:

   • Add variant to LLMProvider enum
   • Implement in LLMService::generate_grounded_answer()
   • Update frontend settings UI

• Ensure Python 3.7+ is installed
• Run python setup_local_embeddings.py to install dependencies
• Check that sentence-transformers is properly installed

• Ensure Ollama is running: ollama serve
• Check available models: ollama list
• Pull a model if needed: ollama pull llama3

• Verify API key is correct and has credits
• Check network connectivity
• Ensure API key has appropriate permissions

• Check console logs for detailed error messages
• Verify all dependencies are installed
• Try restarting the application

• Local Mode: All processing happens on your device when using Local Python + Ollama
• API Keys: Stored locally, never transmitted except to respective services
• Data Storage: All captures stored locally in SQLite database
• No Telemetry: No usage data is collected or transmitted

1. Fork the repository
2. Create a feature branch
3. Make your changes
4. Add tests if applicable
5. Submit a pull request

[Add your license here]

For issues and questions:

• Check the troubleshooting section above
• Review console logs for error details
• Open an issue on GitHub with detailed information

Note: This is an early version. Some features may be experimental or require additional setup.