I am a specialized Audio Engineer providing high-fidelity solutions at the intersection of traditional signal processing and cutting-edge neural architecture. I combine clinical-level attention to detail with an advanced AI toolchain to deliver studio-grade French and multilingual audio content. My workflow is built on a comprehensive, multi-stage signal chain designed for maximum vocal transparency and linguistic accuracy: • Neural Source Separation: Utilizing UVR5 with specialized models (MDX-Net, VR Arch, Demucs, and Roformer) for surgical isolation of vocals and artifact removal. • Vocal Restoration & Enhancement: Advanced deployment of Adobe Podcast AI paired with manual spectral repair, Amplify optimization, and Loudness Normalization. • High-Fidelity Voice Modeling: Expert-level training and inference using RVC (Titan 48k) and SoftVC VITS (Fish SVC) to produce natural, emotion-driven vocal assets. • Precision Engineering: Critical application of Parametric EQ, dynamic range Compression, and Brickwall Limiting to ensure broadcast-standard clarity and LUFS compliance. Whether I am cleaning complex datasets for AI training or mastering a final vocal mix, my goal is to eliminate digital harshness and restore the organic warmth of the human voice.