AI Trainer - Fluent Kazakh Speaker - Portland, US

Role overview This is a remote, task-based evaluator opportunity for fluent Kazakh speakers contributing to the training of large language models. Participants read AI-generated Kazakh text and judge how natural, idiomatic, and culturally appropriate it sounds compared with authentic human writing. Work is paid per assignment rather than as salaried employment, with most tasks requiring roughly one hour of focused effort.

Responsibilities - Carry out side-by-side comparisons of AI-generated Kazakh text snippets to gauge quality and authenticity - Rate AI-written samples for naturalness and idiomatic fluency - Flag tone, phrasing, or vocabulary that feels unnatural or culturally mismatched - Verify that word choice and register suit the intended cultural context - Annotate and tag data with structured feedback on tone, authenticity, and overall quality

Requirements - Native or near-native fluency in written and spoken Kazakh - Strong English skills for following task instructions and reporting observations - A sharp eye for subtle differences in tone, register, and word choice - A computer with reliable internet plus a quiet workspace for uninterrupted reading and writing - A PayPal account to receive task payments

Nice to have - Background in linguistics, translation, or a related discipline - Prior experience evaluating text quality or annotating language data

Benefits and work setup - Typical task pay of around $20–$26 per hour, with rates and durations shown before accepting - Fully remote with flexible scheduling; you choose which tasks to take on - Quick onboarding after passing an initial assessment - Based in or near Portland, though work itself is carried out remotely