The Steel Man and Data AnalysisAs a data scientist, my job is to derive value from data. This can come in many forms; sometimes it looks like training a machine learning model to produce a tool that provides value. Sometimes, it’s analyzing data to derive insights to drive busines...Jul 21, 2025·5 min read
Google Batch for ad hoc parallelization of large workloadsRecently, I worked on a project that involved processing large volumes of 3D wind data across a long time span. The data was segmented into short intervals, and each interval took 8 to 10 minutes to process on a single-threaded instance. Running the ...Nov 29, 2023·3 min read
On Metrics and Mathematical ModelingMetrics are a quantifiable numeric representation that distills complex data down into something more understandable. Businesses use metrics all the time - some common ones include profit, profitability, revenue, cost, churn, conversion rate, etc. Wi...May 1, 2021·4 min read
On the value of preprocessing and feature engineeringWhile I worked in the data science department of a large insurance carrier, the VP of Data Science once said to me, "Everyone has the same math." The intent of this statement was to highlight that everyone gets distracted by sexy data science models....Oct 1, 2020·5 min read
On Agile, collaboration, and the way of workingI recently joined an oil and gas startup which employs the use of neural networks in order to derive rock properties using well logs and seismic data. Joining a startup after working at a large company for the last two years was an interesting transi...Jan 1, 2020·5 min read
On Inherent Obstacles in IoT DataA recent project that I've been working on with my company involves the usage of IoT sensors to collect ergonomics data in order to help investigate products and services that my company as a workers compensation insurance provider can make available...Aug 1, 2019·4 min read
Organized Exploratory Data Analysis using an HTPIn the ever-evolving landscape of data-driven decision-making, harnessing the power of exploratory data analysis (EDA) is becoming an increasingly important tool in every data scientist's toolkit. In a recent project for my employer, a Fortune 500 co...Jun 1, 2019·4 min read
