AI Trainer & Prompt Engineer
Designed and refined over 12,000 prompt-response pairs for GPT-class large language model training. Conducted RLHF evaluations, ranking and rating AI-generated outputs to enhance model alignment with human intent. Audited and contributed to the quality assurance of annotated datasets, ensuring high data integrity and reducing error rates. • Produced prompt-response data for creative writing, research, and factual Q&A. • Maintained output accuracy in the top 10% of contributors. • Collaborated with QA to identify bias and errors in training datasets. • Specialized in RLHF and prompt engineering for language models.