AI Trainer - Kannada - San Antonio

Role overview

This remote, task-based role invites fluent Kannada speakers to evaluate AI-generated text and speech for naturalness, cultural accuracy, and linguistic nuance. The work directly influences how well voice and language models understand and reproduce everyday Kannada. Contributors set their own pace, choosing individual tasks with clear time and pay estimates.

• Compare paired text and voice samples and judge whether the AI output sounds authentic to native Kannada speakers.
• Listen to AI-generated audio clips and rate naturalness, clarity, and vocal quality.
• Flag mispronunciations, unnatural tone, or culturally awkward phrasing.
• Assess whether intonation and pacing match the conversational context.
• Apply structured tags that describe the naturalness, tone, and authenticity of each sample.

• Strong command of natural, everyday Kannada, ideally at native-speaker level.
• Sensitivity to subtle differences in accent, register, and prosody.
• Working English proficiency for following task instructions and writing feedback.
• Sharp attention to detail with the ability to stay focused through uninterrupted review sessions.
• Quiet environment plus good headphones or speakers for audio evaluation.
• Pay Pal account for receiving payment.

• Study or work experience in linguistics, translation, or a related field.

• 100% remote with no fixed hours and no minimum workload.
• Piece-rate compensation, with typical pay around $3.40 per 10-minute task.
• Tasks listed individually so pay, effort, and topic are known up front.