Hulya Alpagu

Data Analyst

June 2024 - Present

The Avery Group, LLC, Remote, United States

Client: U.S. Department of Veterans Affairs

• Enhanced veteran engagement and healthcare utilization through advanced Statistical Modeling and Predictive Analytics, utilizing SAS, R, and T-SQL, resulting in a 15% increase in engagement.
• Designed and automated 15+ Power BI dashboards, leveraging DAX calculations to provide real-time insights, reducing manual reporting time by 40%.
• Analyzed healthcare data using Microsoft Excel and T-SQL, applying advanced functions like pivot tables, Vlookup, Xlookup, macros, and Index /Match for data processing and insights.
• Implemented ETL (Extract, Transform, Load) and ELT (Extract, Load, Transform) processes to integrate and transform healthcare data from multiple sources, ensuring data consistency, accuracy, and operational efficiency.
• Conducted A/B testing and hypothesis testing, driving data-backed decision-making that increased operational efficiency by 25%.

Notable Projects:

→Veteran Home Healthcare Project: Leveraged statistical modeling and data analytics to evaluate veteran engagement and healthcare utilization, achieving a 20% improvement in service delivery through targeted insights.

→Employee Certification Healthcare Project: Enhanced healthcare certification tracking, optimizing data processes and achieving a 30% increase in certification completion, improving compliance and operational efficiency.

Data Science Research Assistant

Nov 2022 - Present

Saint Peter's University, Jersey City, NY

Client: Texas Department of State Health Services

• Managed the end-to-end lifecycle of five Data Science projects, including two in the process of public release, while utilizing Agile Methodology for requirements gathering, planning, design, and deployment.
• Designed and applied Machine Learning models, including Logistic Regression, K-means clustering, Gradient Boosting, and Deep Learning, to analyze Healthcare and Financial data for Predictive Analytics.
• Conducted NLP and Sentiment Analysis using Vader, TextBlob, and BERT to uncover trends in American song lyrics, leveraging Python (Azure) and Big Data techniques.
• Guided and mentored over 50 students through project completion, resulting in a 25% increase in on-time submissions and a 30% improvement in overall project quality

Notable Projects:

→ Interpretable Machine Learning Prediction of Mortality: Developed a gender-specific machine learning model with 95% accuracy to predict all-cause mortality, emphasizing the need for gender-specific care, and won the BAASANA Best Presentation award.

→ Bankruptcy Prediction of Companies (Public in IEEE Xplore): Applied Z-Score and statistical modeling (T-tests, A/B testing) to assess bankruptcy risk in the Financial, Technology, and Pharmacy sectors, achieving 92% accuracy and recommending financial tools for improved forecasting.

→ Analyzing American Song Lyrics (In Progress: Public): Used NLP and Big Data techniques to uncover emotional trends in American song lyrics, revealing an 80% prevalence of negative emotions. These insights informed audience engagement strategies and content development.

Data Scientist

Oct 2016 - Nov 2018

Dogus Group , Istanbul, Turkey

Client: Dogus Automotive, D.ream Group, Dogus Technology

• Built and deployed Predictive Models using Machine Learning algorithms to solve complex busi ness challenges across multiple sectors, including Automotive, Tourism, and Technology, improving decision-making by 20%.
• Utilized Python libraries (Pandas, NumPy, SciPy, Scikit-learn, TensorFlow) for Data Wrangling, Feature Engineering, and Statistical Analysis, extracting actionable insights from diverse datasets.
• Applied Time Series forecasting and ARIMA models to predict trends in automotive and tourism industries, improving resource allocation and decision-making by 15%.
• Created dynamic Data Visualizations and dashboards using Power BI and Matplotlib, enabling real-time data-driven decision-making for stakeholders.

Data Engineer

Aug 2014 - Sept 2016

i2i Systems, Istanbul, Turkey

Client: Turk Telecom, Akbank

• Designed and optimized ETL pipelines using SSIS and Python(Azure) to process large datasets, ensuring high-quality, structured data for analytics and Predictive Modeling.
• Engineered a high-performance Data Warehouse, reducing data retrieval time by 35% and enabling faster Business Intelligence Reporting and advanced analytics.
• Optimized SQL, PL/SQL, and Python scripts, improving Data Processing speed and supporting real-time insights for customer segmentation, fraud detection, and financial forecasting.
• Collaborated with data scientists and analysts, leveraging SQL and Python to prepare clean, struc tured data for ML Models, enhancing Predictive Analytics in Telecom and banking.

M.S. in Data Science

2022 - 2024

Saint Peter's University, Jersey City, NY

GPA: 4.0/4.0

B.S. in Computer Sciencee

2010 - 2015

Kocaeli University, Umuttepe, Turkey