AI Language Trainer & LLM Evaluator
Scale AI
Dec 2024 - Current (1 year 11 months)
Designed and refined prompts and responses in Turkish and English for RLHF and SFT projects. Evaluated AI outputs for linguistic accuracy, logical consistency, and factual correctness. Conducted cross-linguistic analysis, applied NLP techniques to multilingual pipelines, contributed to LLM localization into Turkish, and performed AI red teaming and safety evaluation.