Key Responsibilities
The candidate will be responsible for evaluating generated content across specialized workstreams:
- Conversational Performance: Evaluating natural language interactions against objective quality gates and benchmarking results against industry standards.
- Model Response Evaluation: Assessing response quality to ensure alignment with strict internal standards.
- Visual Interpretation: Testing and validating the ability to interpret and identify visual/image-based inputs with high accuracy.
- Factuality & Verification: Conducting deep-dive audits of text and image claims. This involves itemizing and cross-referencing information.
Candidate Profile & Educational Requirements
- Education:Master's Degree in a relevant field (e.g., Linguistics, Communications, Journalism, Translation & Interpretation, Data Science, or related disciplines).
- Language Proficiency:
- Arabic: Native or near-native proficiency (written and verbal).
- English: High-level (C2) proficiency to navigate internal instructions and communicate with a global team.
- Analytical Skills: Ability to analyze responses for helpfulness, factual accuracy, source reputability, and safety/sensitivity.
- Precision: High attention to detail.
- Efficiency: Ability to meet high-demand, rapid turnaround times.
- Language query vetting workflows in India.
- Candidates must have a master's degree in a relevant field and Native or near native proficiency to handle complex vetting tasks.
- The required candidates should have a master's degree in a relevant field such as linguistics, communication, journalism, translation, interpretation, data science, or related disciplines. Candidates must possess a high level of C2 proficiency in the languages required and be capable of navigating internal instructions, communicating with the global team, and analysing responses for helpfulness, factuality, accuracy, safety, and sensitivity.
- The individuals are expected to have flexibility to work during Indian standard time (IST) hours with some overlap with teams in the Pacific Standard Time (PST) and Central Standard Time (CST) time zones, with a weekly cap of 35 hours.
