5 Opportunities After Data Science: What To Do If Data Science Isn’t A Fit (Part 2/2)
Career Capital is meant to be deployed & as a data scientist you have plenty of it. #recoveringdatascientist
In Part 1 of this two-part series, we answered the following questions:
- ✅ Why Would Data Scientists Leave Data Science? (Or Even Data Entirely?);
- ✅ Some of the traits that make data scientists great a product managers?;
- ✅ What MLOps and data engineers have in common (hint: many of them are #recoveringdatascientists)?
We’re now going to cover another three options data scientists have to capitalize on their experiences and expertise.
Option 3: Utilize experiences & learnings to create products & services for data scientists and data science leaders.
If you understand your customer deeply, you’re only a couple of steps away from creating products and services.
Data science and machine learning products are the new norm and represent a seismic shift in the landscape of technical skills and expertise.
Companies, entrepreneurs, and even technical operators are trying to make sense of the available opportunities.
Engineers, non-engineers, and individuals are trying to ensure they don’t fall behind.
There are plenty of opportunities and whitespace for products and services focused on ML & ML product use cases.
For some ideas, I would highly recommend checking out The Pragmatic Engineer’s 1-person business blog post as well as Swyx’s Part-Time Creator Manifesto blog post.
An example of someone who’s leveraged their experience working as a data scientist as well as data science teams is my friend Vin Vashishta. Vin regularly writes about enabling AI strategy and monetizing AI products at his substack and on LinkedIn.
Option 4: Specialize & Go deeper
Maybe there’s something specific you’re interested & you want to do more of that work.
Data scientists tend to start off as generalists, working on any type of project regardless of whether the project is a time series, computer vision, NLP, recommendation project, etc.
After a certain period, some data scientists find themselves enjoying a very specific area of machine learning or data science and will become “T-shaped” data scientists.
For example, my friend Harpreet Sahota used to be a generalist data scientist (and before that he worked as an actuary) but recently has been deeply focused on computer vision and deep learning.
But finding your “sweet spot” of data science work that is valuable and that you enjoy takes time and producing work.
“Good judgment comes from experience, and a lot of that comes from bad judgment.” (Will Rogers)
Option 5: Switch companies or teams
You would be surprised at how a change in scenery can impact someone’s mood and outlook.
Sometimes the solution to being miserable as a data scientist is to switch teams (especially if your manager or teammates are the main cause) or even companies, especially if the problems are due to culture and inertia.
For example, there were times in my career, especially if big life changes were happening or I was trying to upskill or switch into a new role, I needed stability and work-life balance (and a paycheck) more so than growth opportunities and volatility. A well-established company with a scoped role was ideal.
There were other times when I got frustrated with the pace of change and red tape and a startup environment made more sense. High autonomy and ownership, high pace.
Many data scientists experience pain due to a lack of data maturity and infrastructure, so ensuring that you have the tools and environment necessary to create analyses and models should be at the forefront of your thoughts when interviewing hiring managers.
Final Thoughts on Whether Data Science Is Still A Viable Career Path
So……
This is probably where I should start tossing some numbers about the growth of data science job listings versus other jobs, how economic factors will impact tech industry growth, whether ChatGPT will take over ALL engineering 😱😱😱😱😱😱😱😱😱
… heck, maybe the question that should be asked is whether prompt engineering is the “New Hotness”??!
If you pose the question of whether or not data science is fading fast, you’re guaranteed some robust discussions, similar to the one seen below.
Why The “Data Science is Dead” Debate Keeps Popping Up (While Data Scientists Are Still Being Hired)
And there are a couple of reasons why the debate hasn’t died down and why I believe the data scientist title will stick around for a while.
The existence of research-focused roles where statistical mastery and programming skills are still important
Example: It’s not a simple matter to A/B test email campaigns at digital health companies like Teledoc — because of regulations like HIPAA (Health Insurance Portability and Accountability Act — USA) & concerns around behavioral-based human testing. Health companies typically have fewer users at any point in time than software or consumer product companies like Google, so proper experimentation design (including sample size & power calculations) matters.
The outsized impact of the few research projects that do end up having commercial value
- Think of the current trend of LLM’s and the importance of algorithmic design. When you combine innovation with strategic product design & development, the value of that one project (or portfolio of projects) can make up for years of intensive investment. Most importantly a valuable enough idea can lead to an entire ecosystem of companies, projects, and services built on top of that lower abstraction. But the kernel needs to be developed first.
The usefulness of the “data scientist” title as an abstract class.
- Would the data scientist role still be any more or less confusing with a different name? To be honest, I’m not sure. Although an argument could be made that many data scientist roles (especially product data scientist roles at Big Tech Co.) resemble data analyst roles, I’m not sure we could make the same argument for roles like MLOps Engineer or Data Engineer roles.
- Example: While some data engineers were setting up pipelines and data processing orchestration as data scientists (i.e. the data infra work), my entryway into MLOps through data science was trying to figure out how to productionize a data science model as a live service and how to retrain the model and interpret the predictions. If you had called me a “data engineer” as a data scientist, that would have been more inaccurate than calling me a “data scientist”.
Rather than trying to create a “one-size-fits-all” definition of data science, why don’t we instead treat it as an abstract class with the expectation that each company and team will adapt the role to its particular stack and responsibilities?
The More Things Change, The More They Stay The Same
“It’s tough to make predictions, especially about the future.”
“The future ain’t what it used to be.”
“When you come to a fork in the road…. take it.”
“The future ain’t what it used to be.”
The most successful data scientists, engineers, and content creators understand how luck is an equally important ingredient in a successful career.
However, so is individual initiative and creating opportunities through dedication and hard work.
👉 Did you start learning an important piece of technology right when it was taking off?
👉 Did you spend a significant portion of your career studying mRNA before the COVID-19 pandemic swept the globe?
👉 Did you spend years studying at prestigious universities and working your butt off to become a deep learning expert and AI Engineer at Microsoft right before ChatGPT launched?
Neither you (reader) or I am uniquely gifted in future-scrying but building career capital means never having to worry about a perfect landing in a career or role.
You’ll be prepared no matter what.
Career Capital Is About Having Options To Deploy
And it’s not about getting it right the first time, or the second time, or the n-th time.
Some of my favorite books that I re-read or reference over & over again include:
- Cal Newport’s “So Good They Can’t Ignore You: Why Skills Trump Passion in the Quest for Work You Love*”*,
- Reid Hoffman’s “The Start-Up of You”,
- Meg Jay’s “The Defining Decade: Why Your Twenties Matter And How to Make the Most of Them Now”.
They helped me understand the following about my time as a data scientist (and the career capital I built up during that time):
✅ Career advancement is about the strategic accumulation and leverage of career capital
✅ Careers can be directed and are a result of a combination of engaging in informed risk (in the form of opportunities) and developing a craftsmanship mindset
✅ Much like startups your goal in your career should be to innovate, test, gather feedback and re-adjust
✅ Highly desirable careers are characterized by: Autonomy, Competence, Relatedness
✅ Highly desirable careers are attained through high skill & experience (aka you got to work at it — nothing is handed to you)
And just because you leave data science, doesn’t mean you can’t go back to doing the work of a data scientist.
For example, in one of my prior companies, the Director of Data Science stayed at the company but transitioned into an IC, Principal Data Scientist role.
I know engineers who have hopped back & forth between management and IC roles, while also balancing consulting projects.
My friend Ken also left Data Science at some point and then came back to it. Ken has a lot of projects going on at any point in time but he also still consults as a data scientist while producing content.
And while my friend Nick Wan (as far as I know) hasn’t left and boomeranged back to data science, he does stream at night as a side to his role as the Director of Baseball Analytics for the Cincinnati Reds.
And hey, if none of that works, you could always try becoming an extra in a Youtube influencer’s video or going into DevRel/Solutions Engineering. 🤷🏻♀️😬
Thanks for reading along!
Check out all my other posts about #datascience, #mlops, and #MLengineering here: MLOps by Mikiko Bazeley
Want to connect? Find me at these places 👇
- LinkedIn: https://www.linkedin.com/in/mikikobazeley/
- Youtube: https://www.youtube.com/@Miki_ML
- Twitter: https://twitter.com/BazeleyMikiko
- Mastodon: https://data-folks.masto.host/@mikiko
