Some of the hottest jobs right now are in the fast-growing industry of data science. Though we may not realize it, graduate students in science already have many of the analytic skills necessary to succeed in this field.
David DuPuis, a Huck alumnus, gave a Career Exposure Seminar about his transition from graduate student to data scientist. Dr. DuPuis has a BS in computer science from Washington University in St. Louis and a PhD in neuroscience from Penn State. He has worked as a marketing scientist at MarketShare and is currently a data scientist at ScribbleLive.
Is there anything specific that prepared you for your current career?
My CS background made me a better on-paper fit for a data science role, as did my focus on statistics and methodologies in graduate school.
What are your current roles/responsibilities? How have these changed over time?
Work with product teams to identify potential projects. Scope projects requirements (UI/UX perspective with product; performance/scalability/complexity with engineering). Research potential approaches. Build prototypes of the approaches. Review with product teams. Work with engineering on deploying production systems.
Was this career path something you had always considered?
The career path really didn’t crystallize until right about the time I finished graduate school.
What skills have made you and others in your field successful? Were there any unexpected skills that you needed to learn?
Selecting and starting projects can be surprisingly hard. A good project needs to be valuable and doable. And it may not be evident that a project is going to be fruitful until a lot of time has been invested. Project planning and scoping is done to mitigate this risk as best as possible This is a highly collaborative process between product, engineering, and data science and requires strong communication skills.
What’s the most challenging part of your career?
Right now, the most challenging part of my career is coming up with projects that will be high-impact and can add value to our company’s software offerings. A data science project is only worthwhile if it connects to a product vision. It’s not always clear how to make this happen, which means it can be difficult to scope out projects.
How do you think your career will change in both the near and distant future?
I’m starting to get more involved in product aspects at my company, and I could envision this continuing. Product teams are concerned with the look and feel of the software and how users interact with it. Any data science feature that goes into production will need to be productized. And designing intuitive ways for users to interact with data (and algorithms) has been an interesting challenge.
What can a young scientist do to position him or herself for a career as a data scientist? Any tips on specific ways to network in the field?
Play around with data as much as you can to familiarize yourself with the tools and techniques. Kaggle has some great data sources to do this. Andrew Ng also has a great machine learning class that runs periodically on Coursera. If you can create your own mock project, even better. Data processing, cleaning, and featurizing is a large part of what a data scientist does. Kaggle’s datasets are already clean, so you miss out a bit on this.
After ScribbleLive, where would you like to work?
I’m not sure what’s next or how long I might stay at ScribbleLive. But I want to ensure my next company is analytically sophisticated and has a clear need for someone like me. That may seem obvious, but there is a lot of hype in the field. I want to work for a company that has projects that are worthwhile and will provide value.
How easy/difficult is it to balance work and personal/family life in your career?
I’m consistently working 45 hour weeks and I don’t have a family yet. So right now it’s pretty easy. 🙂
What advice do you have, about anything, for current graduate students?
Career advice:
If you want to work outside of academia, don’t constrain yourself. No one meets all the “requirements” in a job requisition. Find out what’s really expected and try to fill in those holes in creative ways. And you probably have a lot more applicable experience than you realize, such as communication (writing papers, posters, and presentations), leadership (TAing and supervising undergraduates in the lab), project management, and independent learning.
General advice:
Do what you like.
