Michael Kalmus

Data Science | Web Development

Projects

Soundcloud Scraper

Languages: Python, SQL, git
Skills: Web Scraping, Caching
Tools/Libraries: Requests, Selenium, BeautifulSoup4, Plotly, Pandas

Scrapes data from Soundcloud's top charts, stores the data in a database, and creates various visualizations for the user to gain insight.

This Website!

Languages: HTML5, CSS, JavaScript
Skills: Web development and design
Tools/Libraries: Figma, VSCode, Live Server

I built this site from scratch with HTML5, CSS3, and JavaScript. This also was the first project where I used Figma for prototyping.

Twitter Search Engine

Languages: Python, SQL, git
Skills: Information Retrieval, Transformers, NLP
Tools/Libraries: PyTerrier, Pandas

Gathers data using Twitter's API and uses Information Retrieval techniques such as Learning to Rank in order to find the most relevant users for a given query.

Baseball Simulator

Languages: R
Skills: Probability and Statistics, Markov Chain Montecarlo Simulation, Statistical Analysis and Evaluation
Tools/Libraries: tidyverse, ggplot2

Uses play-by-play data for a full season of baseball games to simulate game outcomes, accounting for batting order and batter skill.

Experience

Work

Operations Research / Software Development @ Intel Corporation
(July 2022 - Present)

Automated various procedures with Python, Powershell, and C# and deployed multiple new scripts and processes in production.
(PL/SQL) Decreased size of tables storing historical data by over 80% and improved logic to insert ~1500 rows/minute instead of ~12000 rows/minute.
(Python) Built web application with Flask to convert user input to SQL, query data, display results, and allow the results to be exported or printed.

Graduate Student Instructor @ University of Michigan

Graduate Student Instructor for IOE 373 - Data Processing and Analysis and IOE 333 - Ergonomics
Taught 100+ students SQL, VBA, and Python by creating engaging labs and assignments and walking through labs with an in-person session.

Data Science @ General Motors

(Python) Forecasted incentive take rates using various Machine Learniing techniques and models. Used techniques such as recursive feature selection and models like XGBoost.

Throughput Analytics @ General Motors

Built interactive dashboards with Microsoft Power BI and manipulated data with DAX to format it for visualizations.

Education

Information @ University of Michigan
(M.S.I., 2020 - 2022)

GPA: 4.00/4.00
Concentration in Big Data Analytics

Industrial Engineering @ University of Michigan
(M.S.E., 2019 - 2020)

GPA: 4.00/4.00
Tauber Institute for Global Operations Fellow

Industrial Engineering @ University of Michigan
(B.S.E., 2015 - 2019)

GPA: 3.50/4.00

Contact

LinkedIn

Github

Medium