James Moeller

Data Scientist and Energy Engineer

ATS Automation, Inc.•  May 2015 - Present

Performed Extract, Transform, and Load (ETL) in Python, R, and SkySpark with data sources of live building automation system data, SQL Server databases, Excel files, Access databases, and CSV files. Increased accuracy in reporting building energy usage by performing data cleaning in Python, R, and SkySpark to remove outliers by standard deviation, review of time series plots, and imputation by linear regression. Saved energy with data analysis on HVAC and electrical systems by developing baseline models and suggestions for improvements, estimating gains for improvements, and evaluating performance after implementation. (R, Python, SQL, SkySpark). Performed live monitoring of system performance, automated fault detection, and analytical forecasting through data visualization by creating SkySpark Views, DGLux dashboards, and Power BI dashboards. Reduced time in data analysis for project documentation in AutoCAD with R script in RStudio for performing Extract, Transform, and Load (ETL) to pull device configuration data from text files, format for project documents, and write to an Excel file. Entire sites of over 50 devices were completed in minutes, originally taking 1hr for 3 devices. Saved 6hr/month in report development by creating SQL Server View and importing into Excel with Power Query. Reduced analysis process from 3 months to 3 days per data center with automation of data center alarm data analysis in datasets with 100,000+ rows, by developing Excel VBA macros, leveraging formulas, and performing Supervised Machine Learning to categorize alarms with Support Vector Machines (SVM) in Python. Compressed time in data visualization and reporting from 4hrs/month to minutes by creating SQL Server View imported into Power BI dashboard to monitor data center energy usage. Reduced point name update time from weeks/building to 1hr/building by writing R script in RStudio to dynamically rename BMS points via joining SQL Server tables of device data and object data, and updating with Regular Expressions. Expanded historical energy analysis, originally 1 past month of data, to allow data comparison over years with all available data by leveraging R in RStudio to aggregate archived trend logs in Access with current data in SQL Server database.

Data Center Analyst

RagingWire•  February 2014 - February 2015

Analyzed central plant energy usage for efficiency of operation (Python, R, Tridium/Niagara/Alerton data logging software).

E6 MMN1 Engineering Laboratory Technician

US Navy•  May 2006 - May 2012

Performed detailed chemical analysis and trending on power plant water. Built database for tracking radioactive material generated (MS Access). Supervised 4 person nuclear plant mechanical operation team (leading, mentoring, and growing team).

Western Governors University

Data Science and Analytics, MS•  December 2018 - June 2020

Python, Data mining, SAS, R programming, SQL, Python, Tableau, Statistics. Capstone Project in R and PowerPoint comparing linear regression with Autoregressive Integrated Moving Averages (ARIMA) and Neural Network Autoregression (NNAR), using Principal Component Analysis (PCA) for feature selection, for forecasting building energy usage. Customer Churn project using Principal Component Analysis (PCA)/Multiple Correspondence Analysis (MCA) and Random Forests to predict customer churn. Projects - Link to capstone project files.

Excelsior College

Nuclear Engineering Technology, BS•  September 2012 - October 2013

Java Object Oriented Programming Intro to Statistics and Probability Calculus and Analytical Geometry 3