About this role
Role Overview
Evaluate Kannada AI-generated responses to identify factual errors, reasoning gaps, clarity or tone issues, and other strengths and weaknesses, then produce clear English-language analyses that will be used to help create an ideal AI response later in the project.
Key Responsibilities- Conduct fact-checking using trusted public sources and external tools.
- Generate high-quality human evaluation data by identifying response strengths, areas for improvement, and factual inaccuracies.
- Assess reasoning quality, clarity, tone, and completeness of model responses.
- Ensure model outputs align with expected conversational behavior and system guidelines.
- Write analysis and evaluation artifacts in English.
- Bachelor''s degree.
- Native fluency in Kannada.
- Strong proficiency in English, including excellent writing skills to articulate nuanced feedback.
- Significant experience using large language models and an understanding of how and why people use them.
- Strong attention to detail, able to notice subtle issues others may miss.
- Background or experience in domains requiring structured analytical thinking, for example research, policy, analytics, linguistics, or engineering.
- Prior experience with RLHF, model evaluation, or data annotation work.
- Experience writing or editing high-quality written content.
- Experience comparing multiple outputs and making fine-grained qualitative judgments.
- You identify factual inaccuracies, reasoning errors, and communication gaps in model responses.
- You produce clear, consistent, and reproducible evaluation artifacts.
- Your feedback leads to measurable improvements in response quality and user experience.
Contract work, paid hourly, fully remote and open to candidates globally.
Compensation15 - 20 hourly
Eligibility and Application Process- Fluent language requirement, as stated above: Kannada native fluency and strong English proficiency.
- To be considered for this role you must complete a Bilingual Competency interview in Kannada. This interview is required during the application process and is used to assess Kannada bilingual ability.
- Analyses you produce for the role should be written in English.