- Analyzes and evaluates textual and audiovisual data using 6 different policies for brand safety standards, advertiser suitability, and mis/disinformation across open web and major social media platforms- Collaborates with internal Policy and Data Science teams to reduce policy gaps and areas of confusion to ensure high precision of classifications for improved accuracy and effectiveness of machine learning models- Performs quality assurance reviews across 6 projects and creates training documentation as a subject matter expert for classification alignment across internal trainees and overseas vendor and contractor teams- Leads a web classification project in 8 languages with 21 language specialists to support product expansion and successful market expansion of existing post-bid TikTok products - Created and automated a keyword analysis system to improve accuracy of ontology-based classification of highly sensitive text content in non-English languages by 17%- Pioneered a language-specific keyword-based content filtration process with Google Query language and optimized a SQL query to improve non-English text sample quality- Created training resources for Korean text and video projects on cultural specificities to support high precision and accuracy for classifications by non-native speakers- Spearheaded short-form video classification projects to support the launch of the post-bid Brand Safety and Suitability product in YouTube Shorts

- Communicated students’ academic needs to staff and faculty representatives - Drafted and developed goals during full-day retreat with Pitzer staff, faculty, alumni, and administrators (March 2019)- Classified various academic-related issues at Pitzer into overarching topics to prioritize for the following five years