Native Speaker Recording Task (IA project)

🎙️ Voice Recording Project

We are looking for native speakers of multiple languages to take part in a major AI Voice Recording Project starting in November 2026. This project is part of an international initiative to help improve speech recognition systems and AI language models. Your voice will only be used to train AI systems and never for any other purpose.

💡 Project Overview

Selected participants will record 500 short sentences in their native language. The sentences are simple and designed for Inverse Text Normalization (ITN) training, including numbers, dates, times, addresses, currencies, URLs, and everyday expressions.

🤝 Who is our partner?

We run this project with a vetted American company that serves as a subcontractor for several of the world’s leading AI platforms. They follow strict quality assurance and data-protection standards throughout the project lifecycle.

🧾 Pre-Selection

Before recording, all applicants must complete a short pre-selection form. This form serves as a quality check to ensure:

• You are truly a native speaker of the target language
• Your recording equipment (phone or laptop microphone) and environment are suitable
• Your profile matches the project’s requirements

Why do we ask for details like place/city of birth, age, and sex? To build a balanced and representative dataset, we aim for gender parity and diversity across age groups and regional backgrounds. These details are essential for fair and accurate AI training and help us conduct an objective pre-selection.

Once submitted, your information will be reviewed by our Quality Control Team. If your profile is validated, you’ll be contacted with full instructions for the recording phase.
Click on "Apply" to access the pre-selection form.

🔒 Data Protection & Privacy

Your voice will be used only to train AI speech systems—no commercial reuse, no public sharing, no other purpose. Your personal data will never be sold or shared with third parties and is used exclusively for pre-selection, project administration, and payment. We collect the minimum necessary information and retain it only for the duration of the project, in line with strict privacy practices.

🎙️ Recording

You will record 500 sentences directly from your smartphone or computer using a simple online platform. The estimated time per sentence is about 2 minutes (including reading, brief setup, and quality checks), which totals roughly 1,000 minutes (~16–17 hours) for 500 sentences.

Flexible scheduling: work from anywhere, at your own pace. For example, you can record about 3 hours per day over one week—or choose any schedule that suits you.

💶 Payment

Compensation is based on an hourly rate. For completing all 500 sentences, the typical total ranges between €300 and €600 (depending on language, validated recordings, and actual time spent). Payment is made by bank transfer after your recordings are submitted and validated. Please allow 1 to 5 business days for the transfer to appear in your account.

👥 Requirements

We are looking for 100 native speakers per language. To ensure diversity and linguistic balance, we aim for a 50/50 gender split and participants from all age groups.

🌍 Languages

Arabic (UAE), Arabic (Peninsular), Cantonese (HK), Cantonese (CN), Catalan, Chinese (China), Chinese (Taiwan), Danish, Dutch, French (European), French (Canadian), German, Swiss German, Hebrew, Hindi, Italian, Japanese, Korean, Norwegian, Polish, Portuguese (European), Portuguese (Brazilian), Romanian, Russian, Somali, Spanish (European), Spanish (USA), Swedish, Tagalog, Tamil, Thai, Turkish, Ukrainian, Vietnamese

⏱️ Duration

• Total estimated time: ~16–17 hours for 500 sentences (≈ 2 minutes per sentence)
• Pre-selection phase: open until end of October 2026
• Project start: early November 2026

💻 Equipment

• A smartphone or computer with a working microphone
• A quiet environment with minimal background noise
• Stable internet connection to upload your recordings

🔔 Future Opportunities

This is the first in a series of upcoming AI voice projects. In the near future, we’ll have longer missions and even potential full-time remote roles. Participants from this project will be prioritized for upcoming collaborations.