•   Leverage topic modeling and knowledge graph to enable efficient cross-disciplinary research and new concepts discovery for large-scale datasets (5.4M+ publications) hosted by IEEE Xplore digital library.
•   Create automation to facilitate research collaboration across divisions and increase search efficiency.
•   Design end-to-end ML pipelines to extract scientific concepts, validate through graph linkages; Tag paper with concepts based on relevancy, and establish the parent-child hierarchy between concepts.

•   Mentor graduate-level projects related to data wrangling, modeling, analysis, and application.
•   Capture and solve coding problems in Python, R, Bash, Git for inter-disciplinary students.
Courses:
   •   Fuqua School of Business - Programming for Analysis and Visualization, Data Analytics and Applications course site
   •   MIDS (my program :)) - Practical Data Science course site
   •   Bootcamp - Computational Methods for Social Scientists course site

•   Developed an interactive dashboard on self-defined metrics to visualize the impact of 3D scans of specimens. Derived insights to encourage user contribution and optimize data storage allocations.
•   Remapped user information after database reconsolidation. Prototyped the data analysis pipeline with cleaning and processing.
•   Worked and communicated across functions in a virtual environment setting to drive the product updates.

•   Implemented an advanced CNN architecture Deeplabcut for object detection with annotations to identify key body part coordinates for research subject video data.
•   Performed clustering methods including PCA and tSNE on research subjects' behavior to classify their time-stamped emotional state. Contributed the analysis to the NIH grant proposal and received approval.