Loading...
macOS screen recording with automated PII scrubbing.
Video, audio, clicks, keystrokes, accessibility tree — exported as structured data for AI agent training or your private archive.
Not just video — the full interaction graph. Screen recording, audio, mouse and keyboard events, window changes, accessibility tree state, and browser DOM events. All synchronized to a single timeline.
Microsoft Presidio + spaCy NER detect and redact names, emails, phone numbers, SSNs, and credit cards — in screenshots, database fields, and transcripts. Creates a scrubbed copy. Originals are never touched.
Record system audio and transcribe it with three Whisper backends: faster-whisper and openai-whisper locally, or the OpenAI API. LLM-powered auto-naming uses screen context to label recordings automatically.
Export every event as JSONL — mouse clicks, keystrokes, window changes, accessibility state — with metadata headers. Ready for fine-tuning computer-use agents or building your own analysis pipelines.
Upload scrubbed recordings to GCS and explore them in the web viewer with synchronized video playback, a virtualized event timeline, and system info. Or generate a self-contained HTML viewer for any recording locally.