Transform hours of raw talking head footage into polished scripts in minutes.

RambleAI is an AI-powered desktop application that preprocesses talking head videos, automatically selecting the best clips and reordering them into coherent scripts. Built with Wails, Go, and SvelteKit, it saves content creators 60-80% of their editing time.

RambleAI is NOT a video editor - it's a preprocessing tool that works WITH your favorite editor (Premiere, Final Cut, DaVinci Resolve). It handles the tedious part of editing talking head content: finding the good takes, removing the rambling, and organizing everything into a coherent narrative.

• ⚡ 60-80% Time Savings: What used to take 8 hours now takes 2-3 hours
• 🎯 Smart Clip Selection: AI identifies the best parts of your videos automatically
• 🧠 AI Script Reordering: Transform disorganized clips into coherent scripts
• 🔒 100% Private: All processing happens locally on your machine
• 🔄 Works with Any Editor: Export to your preferred video editing software

• 🎬 Video Management: Import and organize talking head footage
• ✂️ Smart Highlight Extraction: AI-powered clip selection with quality scoring
• 🤖 Flexible AI Integration: Use any AI model via OpenRouter (bring your own API keys)
• 📝 Accurate Transcription: Word-level timestamps for precise editing
• 🎯 Visual Timeline Editor: Drag-to-resize highlight editing
• 📤 Multiple Export Options: Individual clips or stitched compilations
• 🌓 Modern UI: Dark/Light theme with native desktop performance
• 🚀 Fast Processing: Optimized for quick turnaround times

1. Download the latest release from GitHub Releases
2. Install the application:
   • macOS: Open the .dmg and drag RambleAI to Applications
   • Windows: Run the installer
   • Linux: Extract and run the AppImage

1. Launch RambleAI
2. Create a new project
3. Import your talking head video
4. Let AI analyze and transcribe
5. Review AI-suggested clips
6. Reorder clips into your script
7. Export to your video editor

1. 📤 Upload Footage: Import your raw talking head video. AI analyzes speech patterns and content quality.

2. 🎯 Smart Selection: AI identifies the best clips based on clarity, coherence, and narrative value.

3. 🧠 Script Reordering: AI reorders selected clips into a coherent, engaging script using your choice of LLM.

4. 📦 Export & Handoff: Export optimized scripts and clips ready for your favorite video editor.

• macOS 10.15+ / Windows 10+ / Linux (Ubuntu 20.04+)
• 8GB RAM
• 2GB free disk space
• Internet connection (for AI features only)

• 16GB RAM for optimal performance
• SSD for faster video processing
• Dedicated GPU for smoother playback

• Go 1.22+
• Node.js 20+
• pnpm 9+
• FFmpeg (for video processing)
• Wails CLI

# Install Wails CLI
go install github.com/wailsapp/wails/v2/cmd/wails@latest

# Clone the repository
git clone https://github.com/leoaudibert/ramble-ai.git
cd ramble-ai

# Install dependencies
make setup

# Run in development mode
make dev

# Build for production
make build

# Build from scratch (clean build)
make full-build

The project has comprehensive test coverage with multiple testing approaches:

# Run all tests
make test

# Run specific test suites
make test-go          # Go tests only
make test-frontend    # Frontend tests only
make test-coverage    # Generate coverage report

# Watch mode for frontend development
make test-watch

For detailed testing documentation, see docs/TESTING.md.

To test the complete application locally with both PocketBase backend and Wails desktop app:

# Terminal 1: Start PocketBase backend (reset database)
cd pb-be
make pb NUKE=1

# Terminal 2: Start Wails desktop app  
make dev

This setup provides:

• ✅ Clean Database: NUKE=1 resets PocketBase database completely
• ✅ Consistent API Keys: Both services seed the same development API key
• ✅ Remote Backend Mode: Wails app connects to local PocketBase (localhost:8090)
• ✅ Full Feature Testing: Audio transcription, user management, subscriptions

Development API Key: Both services automatically use ra-dev-12345678901234567890123456789012

Ports:

• PocketBase Backend: http://localhost:8090
• SvelteKit Frontend: http://localhost:5174
• Wails Desktop App: Development window

RambleAI/
├── app.go              # Main application logic
├── goapp/              # Go backend modules
│   ├── exports/        # Video export functionality
│   ├── highlights/     # Highlight management
│   ├── projects/       # Project management
│   └── settings/       # Application settings
├── frontend/           # SvelteKit frontend
│   ├── src/
│   │   ├── lib/        # Components and utilities
│   │   └── routes/     # Application pages
│   └── static/         # Static assets
├── ent/                # Database ORM (Ent)
└── Makefile           # Build and development commands

• Backend: Go, Ent ORM, SQLite
• Frontend: SvelteKit (Svelte 5), Tailwind CSS, shadcn-svelte
• Build: Wails, Vite, Earthly
• Testing: Vitest, Go testing, Earthly CI

# Create new entity
make new-entity name=EntityName

# Generate Ent code
make generate

# Reset database (WARNING: deletes all data)
make db-reset

The project uses GitHub Actions with Earthly for continuous integration:

• ✅ Automated testing on every push
• 🔍 Code linting and formatting checks
• 📊 Test coverage reporting
• 🏗️ Multi-platform builds (Linux, macOS, Windows)
• 🔐 Automated macOS code signing and notarization

# Install Earthly
brew install earthly/earthly/earthly

# Run full CI pipeline locally
earthly +ci

# Run specific targets
earthly +test
earthly +lint
earthly +build-frontend

The project automatically signs and notarizes macOS builds using GitHub Actions. This requires the following GitHub secrets/variables:

Set these in your GitHub repository settings under Settings → Secrets and variables → Actions:

• APPLE_DEVELOPER_CERTIFICATE_P12_BASE64: Your Apple Developer certificate exported as .p12 and base64 encoded
• APPLE_DEVELOPER_CERTIFICATE_PASSWORD: Password for your .p12 certificate
• APPLE_ID_PASSWORD: App-specific password for your Apple ID (generate at appleid.apple.com)

Set these as either repository variables or secrets:

• APPLE_ID: Your Apple ID email address
• APPLE_TEAM_ID: Your Apple Developer Team ID (10-character string like ABCDE12345)

1. Export your Apple Developer certificate from Keychain Access as a .p12 file
2. Convert to base64: base64 YourCertificate.p12 | pbcopy
3. Paste the result into the APPLE_DEVELOPER_CERTIFICATE_P12_BASE64 secret

Run this command on macOS to find your Team ID:

security find-identity -v -p codesigning

The signed and notarized app will be available for download as a build artifact and included in GitHub releases.

1. Fork the repository
2. Create your feature branch (git checkout -b feature/amazing-feature)
3. Make your changes and add tests
4. Ensure all tests pass (make test)
5. Commit your changes (commits are validated with conventional format)
6. Push to the branch (git push origin feature/amazing-feature)
7. Open a Pull Request

The project uses Lefthook for automated pre-commit and pre-push checks:

# Install git hooks
lefthook install

This ensures:

• Tests pass before commits
• Code is properly formatted
• Commit messages follow conventional format

RambleAI works with any AI model through OpenRouter, giving you complete flexibility:

• 🔑 Bring Your Own API Keys: No subscription lock-in, control your AI costs
• 🤖 Any LLM Model: Choose from GPT-4, Claude, Llama, or any OpenRouter-supported model
• ⚡ Smart Defaults: Optimized prompts work great out of the box
• 🎯 Customizable: Tweak AI behavior to match your content style

• OpenRouter (recommended - access to 100+ models)
• OpenAI (direct integration)
• Local LLMs (coming soon)

🔒 Your content stays private:

• All video processing happens locally on your machine
• No footage is uploaded to our servers
• AI features use your own API keys
• Complete control over your data

License pending - please check back soon for licensing information.

• Wails - Desktop application framework
• Ent - Entity framework for Go
• SvelteKit - Frontend framework
• shadcn-svelte - UI components
• Earthly - Build automation

For bugs and feature requests, please open an issue.

For questions and discussions, use GitHub Discussions.