top of page

About Me.

I am an astrophysicist turned data science and analytics professional.  During my time in astrophysics, I authored 2 first-author research papers and contributed to multiple others.  My research focuses included how magnetic fields impact star formation in the galaxy, identifying false positives in the search for exoplanets, and researching how false positives could be minimized in future surveys for exoplanets.

​

After receiving my Masters's degree and analyzing future career paths, I decided to move into the data science realm and began working as an Advanced Analyst at Coverys.  There I played an integral role in building out the Business Analytics Department (later renamed to Risk Management & Analytics) from its initial form as a purely reporting department to one utilizing statical analyses, predictive modeling, and extensive data democratization through dashboards.  I then started working at AF Group as a Senior Data Scientist II where I initially helped extend the suite of claims models to also predict litigation outcomes.  I later moved to the Underwriting side to help revamp and complete the enterprise underwriting model project which provided pricing information for the entire book of worker’s compensation business.

 

Outside of work I can be found playing volleyball, refing soccer, tinkering with electronics, and enjoying as many concerts as I can.

Portfolio

Work Experience

2022 - 2024

Senior Data Scientist II

AF Group

Remote

After Coverys, I accepted a position at AF Group working on their Data Science team as a Senior Data Scientist II.  I initially worked on developing predictive models that identified worker's compensation claims that were highly likely to go to various forms of litigation.  I was then transferred to the underwriting team to work on the core pricing model for the Enterprise.  During my time at AF Group I:

  • Developed and pushed to production a suite of models that predicted the chance of worker's compensation claims requiring a full defense or settlement.

  • Was one of the 4 core data scientists that created the Enterprise-wide loss cost model for the entire Worker's Compensation book of business.

  • Tested multiple tools and implemented a new tool for model monitoring to track both underlying data drift as well as potential model degradation.

2015 - 2022

Advanced Analyst

Coverys

Boston, MA

I am extremely proud of the contributions I provided to Coverys and the Business Analytics Department.  I joined shortly after the Department was created and helped develop its capabilities from basic reporting to full predictive modeling.  My primary responsibility initially was to provide analyses of malpractice claims data for internal and external clients.  However, during my time my position evolved to include:

  • Development and management of department and enterprise-wide Tableau dashboards.

  • Introduced automation through R, Alteryx, and Tableau Prep for report creation.  Results included reducing one report's production time from two days down to 30 seconds.

  • Oversaw the adoption of a previously unused Tableau Server instance, increasing usage from 0 to over 270 users within one year.

  • Used my prior research experience to design and implement a peer review process for analytic and modeling projects.

  • Created and maintained data pipelines for enterprise and project-specific data sources.

  • Co-lead on the enterprise predictive modeling initiative which produced the first internal predictive models to gauge insured risk.

  • Evaluated and provided recommendations for potential software solutions and external data sources.

  • Conducted interviews and provided recommendations on candidates for department openings.

  • Helped mentor Northeastern Co-op students.

Additional Experience

2012 - 2015

Graduate Researcher

Penn State Center for Exoplanets and Habitable Worlds

State College, PA

After college, I enrolled in the Ph.D. program at Penn State in Astronomy and Astrophysics.  During my time at Penn State, I worked for Dr. Suvrath Mahadevan in the Center for Exoplanets and Habitable Worlds.  My research centered on the methods used to detect exoplanets, the identification of false positives, and ways to mitigate those false-positive results.  My work and some results included: 

  • Overhauling a software package to automate data cleansing and reduce overhead resulting in a 10x increase in efficiency.

  • Analyzing satellite observations of the sun and Venus to simulate detecting Venus from another solar system.

  • Analyzing satellite observations of the sun to determine what wavelengths have the lowest potential for false-positive detections of exoplanets.  This resulted in my second first-author published research paper.

  • Helped write successful applications for telescope time and additional funding for research projects.

  • Analyzed Sloan Digital Sky Survey observations of Kepler Objects of Interest and identified false-positive results.

2014

Course Instructor

Penn State Department of Astronomy and Astrophysics

State College, PA

I was given the opportunity to serve as an instructor for a Summer 2014 offering of Astro 1: The Astronomical Universe at Penn State's main campus.  In this role I:

  • Created the lesson plan for the course.

  • Developed lecture presentations, homework, and exams for the class.

  • Managed, taught, and accommodated the needs of 26 undergraduate students.

  • Was recognized for receiving excellent feedback for a first-time instructor by the Department Chair.

Education

2012 - 2014

Master of Science Astronomy and Astrophysics

Penn State University

State College, PA

I was admitted to the PhD program at Penn State for Astronomy & Astrophysics in 2012.  I passed both my candidacy exam and successfully defended my 2nd year project for my comprehensive exam.

​

After passing both exams and receiving my M.S., I decided to pursue opportunities in the professional world in lieu of completing my PhD.

2008 - 2012

Bachelor of Arts Astronomy and Physics

Boston University

Boston, MA

Graduated Magna Cum Laude with Distinction in Astronomy and Physics with a minor in Mathematics.

​

I won both the College Prize in Astronomy and the Institute for Astrophysical Research Prize.

Skills & Expertise

Statistical Analysis

Predictive Modeling

Problem Solving

Data Visualization

Data Exploration and Cleaning

Strong Communication Skills

Technical & Scientific Writing

Data Pipeline Creation

Languages

R

SQL

Python

Past Experience:  IDL, HTML, PHP

© 2035 by Robert Marchwinski.
Powered and secured by Wix

bottom of page