Le Scribe Audio is a two-pass AI audio and video transcription SaaS cutting manual drafting time in half (2 to 3 hours of full re-listening per hour of audio instead of 4 to 7 hours of manual typing). Includes interactive playback with global speaker label renaming (diarization), standard 100% editable Word (.docx), PDF, SRT, and VTT exports, an on-demand executive summary followed by a chrono-thematic breakdown, AES-256 file storage in France on OVHcloud with 1-click deletion, Google Cloud Vertex AI inference in Europe (Zero Data Retention), and optional human proofreading at 60 EUR excl. VAT/h by two successive transcribers.