Text Categorization and Sentiment Analysis
Worked on segmentation-based annotation tasks to support large language model (LLM) evaluation projects. Responsibilities included labeling and segmenting text data into meaningful categories (intent, sentiment, factual accuracy, and style). Ensured high precision through multiple QA reviews and consistency checks. Used Argilla to manage large datasets efficiently and applied guidelines to handle ambiguous cases. Contributed to training and fine-tuning LLMs for improved text understanding and generation quality.