High-Quality Voice Data Collection: Hebrew Native Speaker ($600 Milestone)

Overview We are seeking native Hebrew speakers to record high-fidelity speech data for cutting-edge AI development. We are moving away from traditional "voice-over" styles and looking for authentic, expressive, and emotive voices that sound like real people. The Project & Compensation Format: Milestone-based contract. Pay: $600 total upon delivery and approval of 3 hours of audio. Content: A guided mix of scripted personas, semi-scripted scenarios, and unscripted natural speech. We will provide all necessary guidance and prompts. Strict Technical Requirements This project is for AI training; therefore, audio fidelity is our highest priority. We have a zero-tolerance policy for poor recording environments. Professional Hardware: You must use a high-quality XLR or professional-grade USB condenser microphone. Acoustic Treatment: Recordings must be "dry"—no room echo, background hiss, fan noise, or reverb. No Post-Processing: We require raw, unprocessed WAV files. No EQ, compression, or noise gates. Prohibited: No phone mics, laptop mics, or gaming headsets. Who You Are Native Speaker: You must be a native speaker of Hebrew. Please speak in Hebrew for your recording. You can add English if you wish. Expressive & Charismatic: You can flip between different "personas" or moods while remaining grounded and natural. Reliable: You can deliver 3 hours of high-quality audio while maintaining consistent volume and energy levels. How to Apply To be considered, please submit a 1–2 minute voice sample recorded with the exact setup you plan to use for the project. In your sample, please include: A brief intro: Your name, where you are from, and a quick description of your recording gear (Mic/Interface/Software). A "coffee chat": Talk to us naturally about a hobby or passion for 60 seconds. Expressive range: Give us 30 seconds of high-energy or emotive speech. Important: We will be listening specifically for noise floor and room acoustics. If your sample contains background noise or echo, your application will not be considered.