NLP Model Trainer & Data Annotator (College Helpdesk Chatbot)
As an NLP Model Trainer and ML Engineer, I built and annotated training data for a college helpdesk chatbot using Python and NLP methods. I manually defined and classified intent categories, labeling hundreds of student queries to create a structured dataset for intent recognition. Throughout, I evaluated and improved model outputs via iterative feedback and careful annotation refinement. • Constructed the data labeling pipeline: tokenization, text preprocessing, feature extraction, and logistic regression classification. • Labeled and categorized training samples for chatbot intent classification and misclassification analysis. • Iteratively improved data quality through evaluation against ground truth and targeted re-annotation. • Authored technical documentation explaining annotation methodologies, dataset structure, and evaluation metrics.