Regularly, we do a Q&A with Pivotal employees, highlighting expertise from all areas and levels within the company, sharing a front-line perspective on work at Pivotal, and even helping us recruit (like in the area of data science where we are hiring).
In this post, we get to learn from a Q&A with Mariann Micsinai, one of our senior data scientists. With a background in math, financial services, economics, and computational biology, she has worked on some of the most interesting data science projects in the company and solved real-world business problems in the areas of churn prediction, text analytics, financial compliance, and more.
Given it is also International Women’s day, we frame the context of a few of our typical questions to get her perspective on being a successful woman in business and technology.
Could you tell us about how you grew up and got into software?
Well, I grew up in Gyor, a beautiful baroque city in the Northwestern part of Hungary and a 45-minute drive from Vienna. As a wildly curious child, my interests ranged from ballet to astronomy, math and languages, biology and social studies. I was always excited to learn something new about how the world works (my father’s engineering and math influence), enjoy a beautiful painting (my mother introduced me to arts), or study an esoteric language (Finnish, Russian, Norwegian, Tibetan or Chinese anyone?).
My career path tangoed with my interests, so I ended up with a few degrees: a Ph.D. in Computational Biology from NYU and Yale, and a Master’s in Computational Biology, Mathematics, Economics, International Studies, International Relations and Linguistics. One of the common themes has been applied math, and this relates directly to data science.
Would you tell us more about your work background and any of the more interesting experiences?
One cool experience was at a Wall Street litigation economics consulting company. It was influential for me in two regards. One, I learned quantitative techniques to accurately measure economic damages in lawsuits (commercial and personal litigations as well as punitive damages), which made me draw from my background in economics and math. Two, I had to put this complex analysis in simple and easy to understand language for judges and juries. It made me aware of how important it was to be able to tell a complex data story in a simple manner. I love working at the level of interface between different domains, and I sometimes feel like a translator from one science to another.
Some of my more formative years were spent at (the now infamous, rightly or not) Lehman Brothers, where I worked in the Emerging Markets Trading desk in a market risk role. It was an amazing experience, full of complex data (and adrenaline) that was changing in front of me by the New York minute. I had to make sense of it, in its chaotic, continuous stream, and I had to analyze and produce risk metrics on trading activities for a cross-product trading desk, conduct stress tests, and evaluate pricing and risk requirements on newly traded financial products. It was an exciting learning curve for me.
How did you become a data scientist?
After 6 years on Wall Street, I made a 180 degree turn and earned a PhD in Computational Biology, thus getting a little closer to my other roots. While working full time, I studied in the evenings, exploring my interests in applied math. One of the areas that captivated me was statistical genomics, a completely new field with a lot of complex data and problems. One thing led to another, and I received a National Science Foundation grant and started my PhD.
Computational biology was, at that time, an exciting but emerging field—crossing the boundaries of medicine, statistics, computer science, and mathematics. During my doctoral years, my research interests ranged from the analysis of high throughput experimental data to the development of novel computational methods in human cancer genetics. It was an exhilarating time. The field was blooming with more and more data each month while there were rather few analytic tools to deal with it.
It has been great to work as a data scientist. It allows me to use all the parts of my mind as well as the training and experience I have in economics, medicine, finance, and social sciences. My brain gets plenty of regular excercise, and I have never felt more fit!
How did you end up at Pivotal?
I came to Pivotal straight from my PhD program. I actually pushed myself to finish my studies ahead of time in order to join Pivotal. I had an additional fully funded year, but found the Pivotal opportunity very exciting and in line with my background and interests in big data analytics.
One of the main reasons for me to join the Pivotal Data Science team was to be able to continue analyzing diverse, complex, noisy and high-dimensional data. As well, it allows me to work on developing novel computational methods and learn new techniques. Furthermore, I knew the experience would improve and expand upon my technical specialties, which include machine learning, statistics, parallel computing, applied math, risk management, econometrics, human cancer genetics, and next-generation sequencing data analysis.
Most importantly, I knew the work would make an impact in the real world, helping businesses and making a difference to real human beings.
What is exciting about the work you do at Pivotal?
My work at Pivotal mainly concentrates on the financial services industry, and our team also works with many Fortune 500 companies across industries. For data scientists, the work is often cutting edge and uses some of the most powerful and current tools in the industry. I also appreciate the way we work with customers. For me, the ideal customer engagement is one where I can work in close cooperation with subject matter experts and business users. For this very reason, our data science engagements involve multiple feedback sessions with all these stakeholders. The feedback from these sessions helps us to adjust or correct definitions, and often leads to the discovery of additional data sources and insights that can augment the model we are working on. For data scientists, it is a really healthy way to work.
Lastly, I often get to translate our work into articles that explain how certain approaches and capabilities work in the real world. As well, we get to explain the results. For example:
- Churn Prediction in Retail Finance and Asset Management Part 1 and Part 2
- Text Analytics and Natural Language Processing in the Era of Big Data
- 3 Key Capabilities Necessary for Text Analytics & Natural Language Processing in the Era of Big Data
- Financial Compliance: New Frontiers with Data Science
- Re-Architecting Genomics Pipelines to Handle the Rising Wave of Data
Statistically speaking there are not a lot of women in technical roles, yet the Pivotal Data Science team is almost half women. Is there something in data science that you think appeals to women more?
I personally do not think the aptitude for being technical or doing data science has to do with gender.
However, for 9 years, between 2003 and 2012, I taught students in the Economics Department at Barnard College (Columbia University). My courses were highly quantitative in nature, such as Math for Economists, Econometrics, Macroeconomics and Real Business Cycles. During my years of teaching there, I noticed that some women are afraid of quantitative subjects.
Certainly data science involves quantitative subjects, however, the Pivotal Data Science team also involves a lot of creativity (by the way, after mothering so many projects with Pivotal, I am now expecting my first child). It also requires flexibility and being able to transgress barriers. Of course, one has to be a good translator as well—translate business problems into machine learning problems and then translate the mathematical formulations back to insight and value to our clients. It helps to be a good communicator, a conciliator of sciences, and, apparently, knowledgeable across separate domains. Perhaps the variety of skills appeal more to women that are timid to just quantitative topics.
What advice do you have for women looking for a career in data science?
My advice is to move away from gender issues and follow your dreams, interests, and passions. Let others (psychologists, sociologists, and maybe politicians) fret about them. Most of us work in a meritocratic and an increasingly progressive country. We do not have to fight the feminist battles of our grandmothers and deal with the post-feminist resentment that our mothers dealt with. Just be yourself, fully be yourself, and you’ll be ok.
Pivotal is a very modern, leading edge company that very much operates on meritocracy. For the data science team, this kind of work environment has attracted some of the best minds in data science—and half of them just happen to be women.
More About Pivotal Data Science:
About the Author