Aside

Download a PDF of this CV

Contact

Programming

R
Python
SQL
Javascript
Julia

Languages

Spanish
Greek

Disclaimer

Made with the R package pagedown.

References are available upon request.

Last updated on 2022-05-17.

Main

Andrew G. Argeros

Dedicated and driven data scientist with a passion for statistical, analytical, and machine learning approaches to modern issues. Enjoys problem solving through data driven thinking and computational methods.

Skilled in applications of R, SQL, and Python for end-to-end data science. Work has included advanced use NLP, Computer Vision, and tabular methods. Avid presenter at national data science competitions and academic conferences. Looking to futher experience in the corporate sector through leveraging skills in machine learning at scale and applied analytics.

Education

B.S. Computational Data Science; B.B.A. Business Analytics; Minor in Economics

Hamline University

St. Paul, MN

2022 - 2018

  • Advisors: Dr. Stacie Bosley and Dr. Andy Rundquist
  • President’s Scholarship Recipient, Heim Scholar, and 2022 MinneAnalytics Scholar
  • NCAA Varsity Athlete: Men’s Tennis

High School Dipolma

Coon Rapids High School

Coon Rapids, MN

2018 - 2014

  • Graduated with Honors
  • Two time National AP Scholar with Distinction

Research & Teaching Experience

Teaching Assistant: QMBE 3740 - Data Mining

Hamline University

St. Paul, MN

12/2021 - 09/2021

  • Assisting Dr. Brett Devine in teaching 33 students concepts of data science and machine learning in R. Course covers topics such as data quality, supervised regression and classification, and unsupervised clustering and text mining.

Research Assistant to the Dean

Hamline University School of Business

St. Paul, MN

09/2021 - 01/2020

  • Hired in 2020 for ad hoc data science needs in the Hamline Business School. Responsibilites include working closely with Dean McCarthy, Support Staff, and Faculty to effectively manage and deploy analyics and data science projects.
  • Student Analytics Director of DAC @ Hamline high school data analytics competition.

Research Assistant to Dr. Eric Hammer

Hamline University School of Business

St. Paul, MN

05/2020 - 01/2019

  • Analyzed modifications to inputs of Hawk/Dove game theory model through use of agent based simulation modeling.
  • Planned collaboration on a project studying cultures’ proverbs and “pop-culture” on voting behavior. Based on the work of Michalopoulos and Xue (2017) on the effects of folklore on rational voting theory.

Industry Experience

Data Scientist & Software Developer

Shields Health Solutions

Stoughton, MA; Minneapolis, MN

Current - 11/2021

  • Currently directing and implementing at-scale analytics and production grade machine learning systems affecting major health systems, pharmaceutical manufacturers, payers, and pharmacies in the realm of specialty pharmacy. Developed Shields’ data science portfolio, and supporting data science efforts from concept through delivery and maintenance.
  • Building production machine learning systems such as: recommender systems for provider interventions, a tabular model to predict patient risk of non-adherence, and an ensemble-based time-series forecasting API system for members to gauge and predict key performance indicators.
  • Supporting Shields Core Engineering team to build in-house ETL and data software solutions using Python, R, and SQL to create a centralized data warehouse for additional ML capabilities.

Data Science Intern

ExceleraRx LLC - Shields Health Solutions

Minneapolis, MN

11/2021 - 02/2020

  • Supported over 30 team members across all sectors of the business for their needs in predictive modeling and federated analytics. Worked with executive teams across Excelera/Shields to make machine learning a core facet of the Excelera model of operation. Consulted data science teams of member Fortune 100 pharmaceutical manufactuers on machine learning issues. Hired as Full Time Employee in November 2021.
  • Built machine learning systems to identify patients at risk for non-adherence in subpopulations of metastatic breast cancer and hepatitis C patients. Currently in development with several national health systems.
  • Built a production string matching system using Zero Shot Natural Language Processing to match raw prescription text to analyzable data using serverless computing systems.

Consultant Data Scientist

Economic Development Company of Lancaster County

Lancaster, PA

01/2021 - 10/2020

  • Used advanced Natural Languange Processing (NLP) and Computer Vision (CV) methods to analyze real-estate trends within the county. Lead a research project to be presented to Lancaster developers and realtors.
  • Developed a cohort of similar communities to Lancaster, PA using T-Distributed Stochastic Neighbor Embedding (T-SNE) and Density Based Stochastic Clustering (DBSCAN) on Census data.

Consultant Data Scientist

Minnesota Hospital Association

St. Paul, MN

02/2020 - 11/2019

  • Analyzed workforce data on MHA’s members for the assocation’s annual workforce review.
  • Presented analysis to statewide health system leaders.

Financial Planning & Analysis Intern

Northwestern Mutual

Minneapolis, MN

01/2020 - 09/2019

  • Worked on a team of six to analyze, forecast, and manage the financial outlooks of more than two thousand clients across the country. Oversaw client investment processes from onboarding through investment and rebalancing.
  • Used basic forecasting techniques (ARIMA, Exponential Smoothing, etc.) to show trends in portfolio growth, client uptake, and advisor put-through.
  • Built a production invoicing system using R and Shiny to effectively manage the department’s billing and receivables.

Selected Data Science Projects

SimCSE & Image Classification Efficiency

Hamline University Department of Computational Data Science

St. Paul, MN

05/2022

  • Analyzed differences in parameter efficiency, validation accuracy, training times, latency of different image classification deep learning algorithms in PyTorch. Compared Sharpened Cosine Similarity to Convolutional layers for SOTA parameter efficency.
  • Developed a real-time classifier using models in Streamlit. Hosted on Streamlit Cloud.

MLB Team Success: Offense vs. Defense

The Federal Reserve Bank of Minneapolis

Minneapolis, MN

11/2020

  • Analyzed the importence of different statistics on predicting Win/Loss percentage and strategic paradigm shift in MLB. Coauthors Ryan Brauer and Jake Dujmovic.
  • First Prize Winner at Minnesota Economic Association General Conference

Forecasting Soybean Futures: Prophet & VAR

MinneMUDAC 2019

Eden Prairie, MN

11/2019

  • Accurately forecasted the price of three target soybean futures securities. Model comprised of an ensemble of Facebook Prophet and Vector Autoregression. Model Accuracy ~99.5%. Coauthors Lindsey Hawk and Lindsay Steiger.
  • 2nd Place Overall & Analytical Acumen Award Winner
  • Invited to present to industry leaders at FASTCON 2020

The Future of Renewable Energy in New York City

BAC @ MC 2019

New York City, NY

05/2019

  • Optimized and analyzed a solution to convert half of New York state’s energy needs to renewable energy. Coauthors Shanoah Harren, Lindsay Steiger, and Leah Wenner.
  • 4th Place Overall

Publications

Predicting Inactivity in Oncology Patients: Machine Learning Classification in The Excelera Network

White Paper

N/A

03/2022

Dermatology Landscape: Continued Growth Within the Excelera Network

ExceleraRx and ShieldsRx Blogs

N/A

03/2021

  • Coauthor with Angela Ouyang