My Life Practicing “Nomadic Data Science”

I am a data scientist, and I am a nomad, therefore I practice “nomadic data science.” As a data scientist, I do consulting, but I’m also a tech journalist and educator. As a nomad, I don’t have a traditional office so I do my data science at coffeehouses, restaurants, colleges & universities, shared office spaces, libraries, parks, trains, buses; the list is long (one of my favorite destinations for doing machine learning work is the Griffith Observatory here in my hometown of Los Angeles). The only tools I need to practice my craft is a laptop (coupled with Google Drive for extensive cloud storage, and Google Colab for compute resources when needed) and a good wifi connection. Even during times when wifi isn’t available, I have a USB antenna device to use a much slower (yet still useful) broadband connection to the internet.

I enjoy the nomadic lifestyle because aside from the potential isolation working with machines and data all the time, being out in the world allows for a heightened sense of awareness, along side the ability to conduct the scientific method in performing experiments on data. My thought processes seem to be more restricted when enclosed in an office. Conversely, there is a sense of freedom of thought when practicing nomadic data science.

I’m not the only nomadic data scientist

Although I’ve been nomadic for a long time, I have known others who’ve adopted this professional lifestyle. Every once in a while, I’ve noticed what’s on people’s screen at the various venues in which I find myself. Sometimes I’ll see RStudio and some familiar looking R code, or I’ll see a Jupyter notebook with some Python code, or I’ll notice someone running Google Colab. In those cases, I’ll strike up a conversation with my fellow nomad and compare notes about how life on the road is conducive for selecting just the right hyperparameters for a machine learning algorithm.

A while back I got acquainted with my friend Lillian Pierson, an environmental engineer turned data scientist, who took nomadic data science to a whole other level. I’d hear about her wanderlust such as studying for her degree program and writing her popular “Data Science for Dummies” book from exotic island locations. Her practices reinforced my own nomadic slant toward data science where a traditional office environment is not required for success.

Last year one of my favorite local destinations, a very nice West LA Coffee Bean & Tea Leaf store, became a gathering place for other nomadic practitioners including a couple of my ex-students. It was fun to collaborate in an impromptu way. We’d share stories, techniques and strategies for getting the most out of data.

Just after LA got it’s first Philz Coffee over in downtown Santa Monica, I ran over expecting to witness the same tech vibe I appreciated in conjunction with the chain’s Silicon Valley roots, but alas I only found the usual Hollywood screen writers and their very recognizable Word templates. Yes, screen writers are pretty nomadic too!

A Few Caveats

Of course with any alternative work strategy, there are a few caveats. First of concern, being a nomad means you experience many distractions. Over the years I developed the ability to tune just about anything out – loud talkers, music, screaming espresso machines, freaks on the street, crying babies and so much more. Today, others remark how easily I can concentrate in pretty much any environment where the average tech worker might want to run for the hills. This is an acquired ability learned over years time.

Starbucks removes power outlets because they hate the homeless more than they love tech professionals

The essentials for doing nomadic data science are: a clean working environment (sometimes a real challenge in public places), free and fast wifi, and an accessible bathroom is nice. Plus, I’ve been known to stay at a location for several hours, so a power outlet for my laptop is a necessity. Lately, all the local Starbucks stores have removed power outlets (the reason given to me: to stop the homeless from charging their phones) so I’ve been forced to give my repeat business to other coffeehouses in my immediate area such as Coffee Bean & Tea Leaf, Peets, and Philz.

One big caveat is in relation to the current caronavirus global pandemic. Not a good time to be nomadic! I’ve found I had to temporarily curtail my nomadic ways and find a place to hunker down for days on end. This level of sequestering limits my creative juices, but this situation is definitely the exception not the rule. Right now I’m doing an elective “shelter in place” for the time being but for a nomad like myself, it’s weird!

Conclusion

Performing nomadic data science isn’t for everyone, but I believe it can be quite rewarding if given a chance. Late last year, I visited the new Computer Science Department building (Engineering VI) on the UCLA campus (my alma mater) and did some machine learning work in the cool, brand new environment. It was comforting to be back at the place where I witnessed the birth of the internet decades ago. Nomads rule!

Sign up for the free insideAI News newsletter.