Descripción de la oferta
About the Role¿No sabe con seguridad qué habilidades necesitará para esta ocasión? Simplemente lea la descripción completa a continuación para obtener una idea clara de los requisitos del candidato.We are seeking highly analytical and detail-oriented scientific experts to support our AI training and model evaluation initiatives. This role is ideal for PhDs in Physics, Chemistry, Biology, or Computer Science who have a passion for research, critical thinking, and applying domain-specific knowledge to cutting-edge AI applications.You will contribute to the development and improvement of AI systems, including large language models (LLMs) and other machine learning pipelines, by creating, curating, and evaluating scientific datasets, validating model outputs, and providing domain-specific insights.Key ResponsibilitiesCreate, review, or annotate high-quality scientific content and datasets to train or evaluate AI systems.Perform quality assurance on model-generated outputs for scientific accuracy, clarity, and alignment with domain knowledge.Analyze and interpret AI behavior in the context of domain-specific tasks and error patterns.Support the development of guidelines for scientific content generation and annotation.Collaborate with internal engineering, data, and linguistic teams to ensure accuracy and consistency across projects.Conduct domain-specific research and synthesize findings to guide model improvements.Identify and resolve issues related to ambiguity, bias, or misrepresentation in scientific content.QualificationsPhD in Physics, Chemistry, Biology, Computer Science, or a closely related scientific discipline.Strong analytical skills and ability to apply theoretical knowledge to real-world datasets and AI systems.Familiarity with scientific writing standards, peer-reviewed publishing, or lab-based research methodology.Attention to detail and ability to critically evaluate scientific content for accuracy and clarity.Excellent writing, editing, and communication skills.
#J-18808-Ljbffr
