Technical AI Trainer (Specialist)
RLHF
As a Technical AI Trainer (Specialist), I performed Reinforcement Learning from Human Feedback (RLHF) tasks focused on evaluating AI-generated code in Python, Java, and C++. My responsibilities included evaluating and ranking model responses for logical and mathematical correctness and writing definitive ground-truth responses for LLM fine-tuning. I also identified and documented edge cases where models failed in mathematical reasoning or robotic path-planning logic.
• Ranked outputs for mathematical accuracy and security
• Wrote golden responses for supervised fine-tuning (SFT)
• Focused on AI alignment in technical, code-based tasks
• Used proprietary/internal software for code evaluation and prompt response
As a Technical AI Trainer (Specialist), I performed Reinforcement Learning from Human Feedback (RLHF) tasks focused on evaluating AI-generated code in Python, Java, and C++. My responsibilities included evaluating and ranking model responses for logical and mathematical correctness and writing definitive ground-truth responses for LLM fine-tuning. I also identified and documented edge cases where models failed in mathematical reasoning or robotic path-planning logic. • Ranked outputs for mathematical accuracy and security • Wrote golden responses for supervised fine-tuning (SFT) • Focused on AI alignment in technical, code-based tasks • Used proprietary/internal software for code evaluation and prompt response
2024 - Present