In my day-to-day practice, I tackle a dizzying array of technical problems to learn and argue to non-technical managers whether and how data analytics can solve their problems. Technical writing has been critical for me to quickly solve those problems and not be consumed by them, leaving time to think about data strategy, model identification, and other higher value problems.
I have always been a note taker, but my practice has evolved over the past two years, thanks in part to markdown and Obsidian, into a more coherent, linked knowledge base of practical know-how spanning statistical modeling, data engineering, and machine learning engineering.
To posts on this site til, which stands for “Today I Learned…”, are
attempts to make notes to my future self accessible to others who may want
to increase their awareness of opportunities to solve problems they may not
have had a chance to discover yet through professional practice.
Experience is a great teacher, as it forces you to learn and retain information while giving you the motivation to do so. What I write about are solutions to problems I confront in my job. I try not to write about data processing as a generalizable technique, since those are covered in Stack Overflow and elsewhere. In my writings, I expose the constraints involved in implementing a technique and the trade-offs that stem from them.
I welcome feedback, ideas, and opportunities to compare notes and collaborate; message me on Linkedin.
Bio
I began my career as a scientist in the field of environmental geochemistry, working in governmental research settings after getting my PhD. For reasons nearly anyone should be able to imagine, I pivoted to the private sector, taking an MBA to rewire and recalibrate. Steadily moving to the financial sector while teaching myself data science and engineering, how to code, and how to manage those who code, I currently build models on various financial data, along with data ingestion pipelines and validation and reporting modules.