Freelance AI Trainer & Annotator
TextRLHF
Worked as a Freelance AI Trainer and Annotator performing RLHF and response ranking tasks for language models. Responsibilities included model ranking, creation of 'Golden Datasets', and implementation of supervised fine-tuning. Specialized in aligning AI outputs for Egyptian Arabic cultural and linguistic traits.
• Ranked AI model responses based on HHH (Helpful, Honest, Harmless) guidelines.
• Conducted Red Teaming for model safety assessment and bias identification.
• Created and validated prompt-response datasets for fine-tuning objectives.
• Improved factual accuracy, reasoning, and safety via direct annotation and evaluation.
Worked as a Freelance AI Trainer and Annotator performing RLHF and response ranking tasks for language models. Responsibilities included model ranking, creation of 'Golden Datasets', and implementation of supervised fine-tuning. Specialized in aligning AI outputs for Egyptian Arabic cultural and linguistic traits. • Ranked AI model responses based on HHH (Helpful, Honest, Harmless) guidelines. • Conducted Red Teaming for model safety assessment and bias identification. • Created and validated prompt-response datasets for fine-tuning objectives. • Improved factual accuracy, reasoning, and safety via direct annotation and evaluation.
2023 - Present