Shaip AI — AI Evaluation & Audio Specialist (Remote/Contract)
Served as a subject matter expert for audio-based projects by evaluating AI-generated speech for naturalness, prosody, and clarity. Conducted accent and dialect identification by categorizing subtle regional accent differences to ensure diverse training data representation. Evaluated LLM outputs for safety, accuracy, and tone with detailed feedback and participated in high-accuracy transcription of complex audio containing technical jargon, overlapping speakers, and noise. • Audio assessment (naturalness, prosody, clarity) • Accent/dialect identification for diverse representation • RLHF/model output rating with safety and bias reduction feedback • High-accuracy transcription and complex QC/Gold-standard review