Multilingual RLHF and Linguistic Validation for LLM Fine-Tuning
Managed high-complexity data annotation and linguistic validation projects focused on training Large Language Models (LLMs) for the French and Arabic markets. My role involved performing Reinforcement Learning from Human Feedback (RLHF) to evaluate model responses for accuracy, safety, and cultural relevance.