AI Trainer - Kannada - Seattle, US

Role overview This is a remote, task-based opportunity for fluent Kannada speakers to act as evaluators of AI-generated language and voice output. The role focuses on judging naturalness, cultural fit, and linguistic accuracy, then feeding structured feedback back to model developers. Workers choose which tasks to accept and work whenever it suits them, with no minimum hours.

Responsibilities - Compare text and audio samples side by side to gauge how well the AI reproduces Kannada. - Listen to AI-generated voices and rate how natural and convincing they sound. - Identify pronunciation issues, tonal mismatches, and culturally inappropriate choices. - Review intonation, rhythm, and pacing against the conversation's cultural context. - Apply quality tags covering tone, naturalness, and overall authenticity.

Requirements - Native or near-native Kannada fluency with confident everyday usage. - Skilled at catching subtle shifts in accent, tone, and vocal delivery. - Fluent English for reading task briefs and submitting written evaluations. - Consistent attention to detail across many short review sessions. - Access to a quiet space and quality headphones or speakers. - A PayPal account for receiving payments.

Nice to have - Experience or training in linguistics, translation, or a related language discipline.

Benefits and work setup - Fully remote, flexible scheduling with no minimum hours. - Piece-rate pay averaging roughly $3.40 per 10-minute task. - Individual task postings show pay, time, and topic before commitment.