Kelsey H. Email and Phone Number

• Leveraged Star schema data model to design and develop a relational database on Snowflake, integrating automated data generation function with 20+ user-defined statistical characteristics to provide synthetic data for Engineering professors on various research projects.• Initiated an auto-grading project for predictive modeling assignments, utilizing statsmodels, scikit-learn and other Python packages to conduct automatic comparisons between students' input and solutions, achieving 89% accuracy while reducing 40% time spent in manual labor.

• Leveraged python for patient arrival dataset cleaning and preprocessing, identifying critical patterns in patient flow.• Constructed an interactive Tableau dashboard updated on a daily basis to monitor trends in patient type and length of stay, reducing 15% in time spent in reporting.• Developed a flow chart through Visio and employed simulation software to dynamically simulate patient queuing process, optimize beds resource, reducing 30% patient waiting times while maintaining a 90% bed utilization rate.

• Preprocessed noisy manufacturing process data containing 290M+ rows using PySpark, addressed outliers, applied statistical testing, and transformed the large-scale data to internal database to build 3 master tables. • Developed K-Means clustering and KNN model for mobile/fix device classification, leveraging external MAC address manufacturer lookup API for validation, refined the model with an accuracy of 85%.• Implemented Gradient Boosted Tree (GBT) model to predict future machine failure occurrences, applied advanced feature engineering and temporal labeling, enhancing proactive maintenance planning efficiency by 25%.• Architected an auto-detecting system dedicated to protecting company’s valuable assets, integrated Twilio API for real-time alerts via text message, improving security efficiency by 20%.

• Designed adaptive product purchasing surveys and performed A/B testing to identify customer preferences.• Initiated focus group to investigate farmers’ adoption patterns of low-carbon agricultural techniques, directed field research on the "Climate-Smart Agriculture" pilot to obtain valuable real-world insights.• Developed a probit regression model using Stata to analyze data collected from 116 surveys to discern farmers’ perceptions regarding the influence of carbon emissions on the adoption of pro-environmental and circular agricultural technologies.