Generative AI Content Specialist | Reviewer (Oracle Level)
Outlier AI
Jan 2025 - Current (1 year 6 months)
• Safety & Ethics Evaluation: Evaluated AI-generated outputs for potential bias, harmful content, and safety violations, ensuring compliance with strict ethical guidelines.
• Pangolin Translations SFT Project: Performed safety-focused review and adaptation of localized content across Ukrainian, Russian, and English to maintain cultural sensitivity.
• Pogo and Elbow Man Projects: Authored diverse prompts and defined specific evaluation criteria to test model robustness and adherence to instructions.
• Cypher Project: Conducted side-by-side model evaluations (Ranking), providing high-fidelity feedback to improve the truthfulness and safety of LLMs.
• Aether Project: Worked across various domains, including visual evaluation, factual verificat
