Even though my undergraduate degree is in electrical engineering, I never had the professional title of Electrical Engineer. I did some electrical engineering in my first job as a Hardware Test Engineer and used those skills some at MTS Systems, but my career pretty quickly become software focused. My software titles have included Software Engineer, Engineering Lead, Engineering Manager, Test Manager, Test Architect, and Architect. I’ve never held the titles of Program Manager or Systems Engineer, but these would have been good descriptions of my role at different points in my career.
Of course, describing what you do is a primary function of a title. This is why I have some misgivings about my current title of Software Architect. It’s a pretty nebulous title in general as the architect role can vary significantly in the mix of requirements, design, and hands on development. What I’ve done over the last few years is “lead broad technology initiatives across a multi-site team and ecosystem”. I guess this isn’t that far off the Wikipedia definition of a “software expert who makes high-level design choices and dictates technical standards, including software coding standards, tools, and platforms”, but that is also pretty broad.
My role is evolving once again – something that I am constantly striving to do in my career. Over the last few months, my job has shifted from part-time analytics focused to primarily analytics focused – Power BI + data architecture + metrics definition + telemetry exploring + ML + enabling data driven decision making across the org. Also contributing this transition is the fact that I handed in the last of my coursework for my MS in Data Science yesterday (woo hoo!). Is it time to change that nebulous software architect title to data scientist?
That’s a harder question than one might think as the data science ecosystem is going through some challenges and growth with titles and roles. Here’s a sampling:
- The Forbes article, Radical Change Is Coming To Data Science Jobs, speculates that data scientists will be split into five types of specialists within a few years.
- A couple of podcasts have explored this topic in the last month – Linear Digressions in their The Great Data Science Specialist vs. Generalist Debate episode and Not So Standard Deviations in 78 – A Sheep in Every Kitchen.
- The HBR article, Why Data Science Teams Need Generalists Not Specialists, makes several arguments against specialization.
Similar to architect, I have some misgivings about calling myself a scientist. This is probably the pragmatic side of my engineering background showing through. There are some engineering titles that appear in the articles and discussions: Data Engineer and Machine Learning Engineer are a couple of examples. But these are titles for specialized roles that my generalist role doesn’t map to.
I’ll resolve this dilemma by going back to the definition that I highlighted in one of my first posts, Why Data Science?: “Data science is the transformation of data using mathematics and statistics into valuable insights, decisions, and products” (from Data Smart: Using Data Science to Transform Information into Insight, by John W. Foreman (http://www.john-foreman.com/data-smart-book.htm). This broad definition encompasses descriptive stats, data architecture, data wrangling, big data processing, visualization, story telling, predictive analytics, prescriptive analytics, and more. The breadth of the work combined with the focus on driving insights, decisions, and products is why I’ve been pursuing a career transition to data science.
I’m going to go with “Applied data science” in my LinkedIn profile. Although I can’t figure out how to work engineer into the title, I somehow feel better with the adjective ‘applied’ and the use of ‘science’ instead of ‘scientist’. I expect the fledgling data science field will continue to evolve, as I will. So I expect this title to change. But regardless of the title, I’m excited about the work and the opportunities in data science!
Picture details: Ice Out Day, 4/23/2019, Canon PowerShot SD4000 IS, f/5, 1/10 s, ISO-1600, –1 step.