Raleigh-Durham, North Carolina Area
• Implemented a natural language processing, named-entity recognition (NLP, NER) project from start to finish to enhance automated document translation, expected to be included in December 2019 production release.
• Researched dozens of NLP libraries and experimented with NER capabilities of Flair, spaCy, and Amazon Comprehend.
• Trained over ten spaCy NER models achieving ~93% F1-score after creating 2000 annotations, then performed elementary data analysis on results to offer data-driven suggestions for future work.
• Illustrated models’ performance in final presentation by a building a full-stack web application visualization tool using a React.js frontend and python Flask backend.