Recently, I have been interviewed for an intern newsletter at work. Beyond usual questions and the fact that Moneyball is not known by everyone, I was wondering why everybody thinks we are "analysts".
It's true that, for ease of language, we say we "analyze" data.
In fact, that's the work of data analysts. Data scientists and engineers analyze data but not the way we intend to.
The problem has at least two faces:
lack of data literacy everywhere. Even in IT positions. No shortcuts: education and time will be needed.
Misuse of position terms. The data-scientist, data-analyst, data-engineer triptych is not that relevant those days.
Within small companies but large scope, the term "Full Stack Data Scientist" could do the trick to better appreciate the skillsets of those who try to ingest, transform, model, analyze and show data in their daily work1.
Machine Learning Engineer, Data Ops, Analytics Engineer, Business Analyst are much more accurate words to describe the large landscape of data positions.
Overlapping job names come from the nature of the data lifecycle and early domain maturity. Like in software, we could see more scoped statutes in the future like "Tensorflow Engineer", "Tableau Developer", "Airflow Engineer" or more explicit positions like "Monitoring Analyst".
Expected Contents
Tidytuesday aficionados
Tidytuesday is one of the most famous code challenges in data visualization. Here are two repositories with awesome creations and corresponding code.
StackOverflow for teams
We use StackOverflow at least one time every day. Why not using the same platform within your organization for better onboarding, problem tracking, recurrent questions, etc…?
StackOverflow has a product for this.
Would love to see it in action. Maybe using Github Issues as a proxy could do the trick?
Deep thinking on AI and how our brain model reality
Artificial intelligence is still artificial.
This article explains really well how getting to a level of resiliency, consistency, and flexibility similar to the one we have as human beings will be possible with a combination of rich versatile models and voting systems.
Colors in action
I am collecting color tools for data visualization… Discovering one every week or so. VizPalette is built by former Apple/Netflix data visualization guru Elijah Meeks and his friend Susie Lu (also from Netflix).
It’s simple but pretty efficient to find a good color set for your plots.
Shopify CEO interview
Probably my best 2021 blog post read for the moment: the interview of Tobi Lukte, Shopify CEO.
Beyond the actual interview which is a bit lengthy but very (very) interesting, other discussions from The Observer Effect interviews are very appealing too.
Looking forward to reading them…
The Blog Post
Event Driven Architecture For Business.
I recently discovered that what I was writing in my last piece “You don’t need an orchestrator” is called event-driven architecture. Late to the party.
Why not using the same kind of approach for business decisions?
Beyond The Bracket
Getting a lot of “wow” effect when scrolling my Behance feed, looking at these incredible digital illustrations. It’s not very concrete, but I love to build my ideas from disparate areas.
The work of Filip Holdas is stunning. His series on “Cartoon Fossils” is very original and funny. It inspires me how “wow” effects can come from simple ideas.
The greatest teacher is called “doing”.
With summer and some (good) life changes coming I’m looking forward to digest those last six months and learn new things after the pool break :)
I got discussions with some of you about featuring articles in The Blog Post section. Drop me a message if you have a blog post idea or tutorial you would like to write. Would love to give some help on it :)
Anywhere you are, enjoy the sun! See you next month.
The way football clubs are filling data positions is very similar to small companies where a “data guy” has to know about everything from infrastructure to business… Seems like stakeholders’ maturity is coming late.