Nathan Sanders

Sector: Industry

Field: Business

Occupation: Data Scientist

In this Paths Through Science profile, we talk to Nathan Sanders, chief scientist at Warner Media Applied Analytics. Sanders leads a team of physical, social, and computational scientists engaged in performing experiments, developing novel statistical and machine learning models, generating remarkable insights about consumers and content for the global media and entertainment business. Nathan talks with us about his path from completing Ph.D. in astronomy, to building a career in data science. You can check out other Paths Through Science profiles here, and if you’re attending Fall Meeting, consider attending our Paths Through Science, Live! event.

What inspired you to become a geoscientist?

As an undergraduate, I had the opportunity to work with samples from the NASA Stardust mission, doing microscopy and image analysis on commentary particles. Looking at primordial material from our solar system, returned from endless tranquility in space to the Earth’s surface by a technological marvel, at hundreds of thousands of times magnification in real time using a transmission electron microscope was truly awe inspiring. As an undergraduate, one of my most formative experiences was volunteering with the Michigan State University Science Theatre, a student-led organization that performs live science demonstrations across the state for schools and public events. Interacting with students and the general public through this work made me realize how excited and passionate I am about scientific concepts. Inspired by these experiences, I went on to complete a PhD on statistical modeling of photometric and spectroscopic observations of core-collapse supernovae (the explosions of massive stars).

What skills have been critical to your success?

I believe that communication is the single most important skill that scientists need to succeed in their work. The ability to translate scientific research is critical in the real world. While its not always recognized and valued for its immense importance, it may well be what determines whether you get the job after your next interview, or whether you receive the grant you apply for. After all, the only value your work will have in the world is the value that you can succeed in communicating. Even the most rigorous, insightful, and novel scientific research will be wasted if you cannot convince others that it is important and relevant to them.

What technical skills have helped you in your career path in data science?

I have found familiarity with probabilistic statistical modeling to be invaluable for my work. Probabilistic (Bayesian) models apply a mathematical framework to explain natural phenomena, applicable to anything from estimating the underlying spatial distribution of chemical components in soils to measuring the masses of exploding stars. Having the dexterity to design original models tailor made for the problems you encounter in your work, guided by your domain knowledge and understanding of the theory of your discipline, is empowering. Even when using off the shelf software and long-established models, or when doing analysis from alternative perspectives, a grounding in probabilistic modeling is beneficial. These models help scientists understand the nature and limitations of their findings.

What are your favorite aspects of your current job?

I really enjoy working with brilliant colleagues from all sorts of different scientific, engineering, and business backgrounds. They bring different perspectives to the challenges that face us in the entertainment industry. I think that a diversity of perspectives is instrumental to the approach we’ve developed, and makes for a really dynamic work environment where my colleagues and I are constantly learning from each other.

How have your mentors shaped your career path?

I’ll always be especially grateful for my PhD supervisor, Alicia Soderberg, and other mentors in my department, like Alyssa Goodman and Avi Loeb, who encouraged me to explore my interests in other fields. While I was a graduate student, I worked with other students to found Astrobites (a science writing collaborative) and ComSciCon (a science communication conference series); I added computational science as a secondary field to my degree; I did a science policy fellowship at the Massachusetts State House; and I volunteered at nonprofits related to science education and environmental science. In retrospect, it’s crystal clear to me that these experiences made me a better Ph.D. student. They made me a more effective collaborator, they made me a more efficient data analyst, they made my publications stronger, and they connected me to a broader scientific community that provided research collaborations and mentorship. I hope that everyone who mentors students recognizes that while we all have a finite amount of time in our day, the way we spend that time is not zero-sum; activities outside the lab can improve and accelerate your work inside of it and help students prepare for diverse careers.