Le Scribe Audio

Le Scribe Audio (published in France by LE SCRIBE AUDIO EURL, SIRET 89004992700043, Le Rheu — Statutory Manager: Mme Miora HARILALANJATOVO; submitted by Jérémie TERRAL — Assistant de direction, LE SCRIBE AUDIO) is a professional two-pass AI audio and video transcription SaaS engineered for French and European organizations, works councils (CSE and CSSCT), local governments, public institutions, academic researchers, legal teams, journalists, and professional transcriptionists.

Whereas typing a full verbatim or clean-read transcript manually from scratch typically requires four to seven hours of focused headphone work per recorded audio hour, Le Scribe Audio's two-pass AI pre-transcription workflow cuts total drafting and full re-listening review time in half, down to roughly two to three hours of human review and final formatting per hour of audio.

Key Capabilities and Editorial Workflow

• Two-Pass AI Speech-to-Text Engine: Combines acoustic speech recognition with contextual linguistic refinement for multi-speaker debates, interviews, municipal council sessions, and technical hearings across 99+ languages.
• Interactive Synced Audio Player & Speaker Diarization Coding: Click any word to jump directly to the exact audio timestamp, adjust playback speed, edit text live, and globally rename or code speaker diarization labels (for instance replacing generic speaker labels with M. Martin, Mme Dubois, or Enquêté P1 across all turns) prior to export.
• Standard 100% Editable Exports: Export transcripts as standard, 100% editable Microsoft Word (.docx) documents ready to customize to your organization's layout after review, alongside PDF, plain text (.txt), and timestamped SRT/VTT video subtitles.
• On-Demand Structured AI Summary: Generate in one click an executive summary followed by a detailed chrono-thematic breakdown of the recording.
• Optional Double Human Proofreading: When certified human verification is needed, order full re-listening and correction by two successive human transcribers at €60 excl. VAT per recorded hour (€1.00 excl. VAT/minute).

European Hosting & Data Protection

Audio files and transcripts are stored in France on OVHcloud infrastructure encrypted with AES-256, with instant one-click permanent deletion controlled by the user. AI inference is processed via Google Cloud Vertex AI in the Europe region under strict contractual Zero Data Retention (no data retention and zero model training on customer recordings). Every new account includes 60 free transcription minutes without a credit card.