Malicious URL and Phishing Email Classification
Annotated a dataset of 7,000+ emails and URLs to train a threat detection model. Tasks included labeling items as "Benign," "Spam," or "Malicious" and extracting key entities like suspicious IP addresses and sender domains.