AI Training Specialist (Contract) — Outlier — Scale AI
Scale AI
Document
RLHFWorked as an AI training specialist on large-scale LLM quality and alignment tasks using structured human feedback. Designed and applied evaluation rubrics to score and rank model outputs, then provided feedback to inform fine-tuning and alignment decisions for frontier language models. Curated and annotated domain-specific datasets to strengthen training corpora for computer engineering and embedded systems contexts. • Developed multi-criteria rating rubrics for accuracy, coherence, tone, and instruction fidelity • Performed human feedback collection via comparative ranking and output scoring • Engineered structured input schemas and output validation criteria • Maintained quality ratings above platform benchmarks by documenting edge cases and failure modes