About this role
Overview Hebrew exposes AI through grammatical gender: verbs and adjectives agree with the person addressed, and the model routinely gets it wrong. It also blurs modern usage with biblical register and struggles with the root patterns a native writer navigates by instinct. As a Hebrew Expert, you'll judge and correct AI output in Hebrew, teaching models to keep gender and register right. What you’ll actually do - Review translations between English and Hebrew for meaning, flow, and correct gender agreement. - Rate and rank model output in Hebrew, catching mismatched gender and phrasing that drifts into the wrong register. - Write the better version when the model misses, natural enough to read as native. Roles this fits Common backgrounds: Hebrew Linguist, Bilingual Reviewer, Localization Specialist. What we look for - Native or near-native fluency in Hebrew. - Attention to gender agreement and to modern versus formal register. - Clear written English: your explanations are the training signal. - No degree required. We care about what you can do, not where you learned it. How it works Apply Qualify Work & get paid Compensation Up to $40/hr depending on task difficulty and specialization. Many contributors add $100k–$200k a year; some make it their full-time income. About DataAnnotation DataAnnotation is where 1M+ experts train the world’s leading AI models. Hundreds of millions paid to contributors to date, and the average contributor stays 5+ years. Flexible, remote, and always project-available.