Why This Chief Data Officer Sees Failure as a Learning Cycle
Western Digital is pleased to be a supporter of the Women in Data Science (WiDS) Podcast. This new series features women leaders across the data science profession, as they share their advice, career highlights, and lessons learned along the way.
How did being raised in a family of mathematicians prepare Janet George for being the Chief Data Officer/Scientist of one of the largest data infrastructure companies in the world?
Janet answered this question and many others in a wide-ranging interview on a newly launched data science podcast. Entering her fourth year as Fellow and Chief Data Officer at Western Digital, Janet is both an advocate and practitioner of data science throughout the business, from manufacturing to security to product development and more.
In this article, we highlight five sound bites from Janet’s recent interview. To hear the full interview, tune in to the podcast interview below!
1. What sort of advice would you have for a woman or man who is listening right now and thinking, “I would love to [try new things] and learn, but I’m too afraid of taking the plunge.”
JANET: I would say that it’s okay to fail because there are lessons in failing. The lightbulb was invented because someone failed 99% of the time and then that 1% of the time they succeeded, right? So, I don’t look at failure as “failure”. I look at failure as a learning cycle. I set out with an intention to say “What will I learn in my first learning cycle?”—which could just be failure. So, I think you have to fail fast and fail quickly, so you can advance toward your second learning cycle. If you treat failure as simply a series of learning cycles, you won’t look at failure the same way as you look at it if you think of it as just failure.
2. What has it been like for you as a woman, making your way up to this level of Chief Data Officer? And what sort of advice would you have for aspiring young women who are listening to you and saying, “That’s me 30 years from now.”
JANET: I think there’s so few women in the scientific fields. And, so, to have a woman represented in the executive team or at the table in the scientific fields—bringing deep technical, business, and strategic value to the company—is very rare. There’s a huge gap. I often find myself as the only woman in many of the executive meetings and leadership meetings. I think it’s a disservice to society as a whole if we don’t bridge this gap very quickly and have more women in executive seats and in executive positions.
That said, how do I influence when I have only one voice at that table? That is again a skill that you hone. I think that you can be very, very clear and intentional with what you’re saying. And you can be clear about leading. I’ve always led by example. Examples of execution, examples of leadership, examples of strategy, so on and so forth. There’s a great deal of influence involved.
You have to build relationships. You have to influence at the highest levels. And you have to be very credible in the company…. I form a lot of advocacy and I form a lot of allies in the executive team. I don’t fight battles that are negative. I fight battles that are positive and lead somewhere.
Read how six WiDS Ambassadors from around the world used data science to bring their communities together.
3. You must have had a lot of courage to jump into this position because you weren’t really familiar with this area. What was that like for you?
JANET: Believe it or not, I thought that it would be interesting because it was a big learning curve for me on the device side and domain side. On the computational side, it was actually a comfort zone for me. The science was comfortable; I’d already done that. I’d already done scale. The domain was different. And, so, I had to learn deeply about the device physics domain. I started learning intimate things about failed bit counts, failure rates, temperate testing, how voltage affects different cells and different rates… The whole manufacturing world is very fascinating to me and I’m continuously learning in this field.
4. What part of mathematics has been really useful for you in your current job?
JANET: Every aspect of mathematics, especially linear algebra, has played a very significant role. So, when you think about the computations of scale, genetic algorithms, and its applications, regression type algorithms, or even neural networks, it’s computationally heavy, it’s mathematically heavy. How do you calculate the cost functions? How do you converge very quickly to where the error mechanisms are or where the failure data is? And how do you go to what is important? What is critical? What you should care about from a business value?
5. What do you like best about your position as Chief Data Officer and Scientist?
JANET: How do you go from tens of thousands of parameters to the 25 or 30 most-critical parameters? This is where machine learning and AI come into play. And this is the part of my job that I love very, very much… When Big Data meets manufacturing, that’s what it’s about. So, the scale of data that I get to play with is the part that I’m most excited about.