Founder & AI Product Lead — RLHF & LLM Data Evaluation
As Founder & AI Product Lead, I led the evaluation of AI model outputs and designed prompt-based data for reinforcement learning via human feedback. I tested and validated LLM responses, mapped failure modes, and assessed annotation quality’s effect on AI performance. This experience included extensive hands-on review and feedback directly applicable to RLHF and LLM data workflows. • Designed, annotated, and rated LLM response datasets for automation products • Conducted prompt evaluation and managed guideline-driven RLHF workflows • Documented annotation edge cases and quality impacts for model optimization • Used LLM and chat interfaces for evaluation and feedback cycles