Freelance Linguistic Annotator (Speech annotation, transcription QA, LLM output evaluation)
Freelance work involved creating and reviewing speech-related annotations for AI training data. The role included transcription QA and evaluation of language model outputs for FR/EN datasets from external vendors. Work focused on speech content and quality, along with semantic ratings and instruction-following signals for model training and assessment. • Transcription review and quality control for multilingual audio data • Evaluation of LLM outputs including semantic and instruction-following aspects • Annotation guideline adherence and structured QA workflows • Support for AI training data vendor deliverables