Al Data Annotation & LLM Evaluation specialist
Worked as an AI Trainer contributing to the development and optimization of Large Language Models (LLMs) through high-quality data annotation, evaluation, and feedback processes. Scope of Work: • Created, reviewed, and quality-assured text-based training datasets for LLM fine-tuning • Annotated and categorized text data using structured metadata frameworks • Performed Named Entity Recognition (NER) and text classification tasks • Evaluated AI-generated outputs for accuracy, coherence, safety, and relevance • Conducted RLHF-style evaluations to improve model alignment and response quality • Wrote and refined prompts and responses for supervised fine-tuning (SFT) tasks • Translated and localized content into Danish while ensuring linguistic and cultural precision • Performed red teaming tasks to identify weaknesses, bias, hallucinations, and safety risks • Tested AI systems and documented errors, inconsistencies, and improvement areas