AI Tutor / ML Engineer at Outlier AI

Fine-tuned large language models using RLHF, improving reasoning accuracy by 15% on benchmark datasets for higher-quality ML system optimization. Reviewed and scored 1,000+ AI responses per week using rigorous statistical evaluation to reduce factual errors by 20% and improve reliability at scale. Designed and validated 50+ prompt engineering templates that reduced debugging time by 25% through iterative analysis and problem-solving.