Work from Home Linguist

Job Responsibilities: Apply linguistic expertise in syntax, semantics, pragmatics, and sociolinguistics to support LLM and generative AI Systems Collaborate with linguists, data operations teams, and ML engineers on data collection, curation, annotation, and localization efforts for model training and fine-tuning Contribute to the development and maintenance of annotation schemas and guidelines for LLM training data (e.g., instruction-tuning, preference labeling, RLHF) Evaluate and quality-check datasets used for pre-training, fine-tuning, and alignment of language models Support the development of programmatic methods for generating synthetic annotated data at scale Assist in model evaluation efforts including prompt-based testing, red-teaming, and linguistic error analysis Contribute to AI safety and responsible AI practices through linguistic review of model outputs (e.g., hallucination detection, bias identification, tone and pragmatic appropriateness) Participate in experiments to assess data quality, annotation consistency, and downstream model performance

Required Skills: Native or near-native fluency in English and at least one additional language Knowledge of syntax, semantics, pragmatics, sociolinguistics, corpus linguistics, and other areas of linguistics Familiarity with Large Language Models (LLMs), their applications and data practices (training data, evaluation, prompting, fine-tuning) Exposure to LLM evaluation methodologies (human evaluation, automated metrics, adversarial testing) Experience working with semantic ontologies, taxonomies, or intent/slot frameworks Proficiency using AI Agents/Chatbots Experience with database queries and data analysis processes (SQL, Python, spreadsheets, R, Unix, or others) Experience working with speech and text data in multiple languages Comfortable working in a fast-paced, highly collaborative environment with evolving priorities Nice-to-have Skills Master's degree in Linguistics, Computational Linguistics, Language Technologies, or a related field Familiarity with machine learning frameworks, NLP libraries, and tools (e.g., Hugging Face, spaCy, NLTK, PyTorch) Exposure to statistical language modeling or training data pipelines