Carolyn Phillips, 39, Lead Data Scientist at Neurensic
Neurensic uses machine learning — advanced mathematical techniques — to detect when traders are doing fraudulent, manipulative things in the financial markets.
We started in 2015. We have about 25 people. There are four of us on the data-science team. I develop a lot of the algorithms that we use to find the manipulative patterns of behavior that are embedded in large data sets.
We’ve got guys who’ve been in different parts of the industry for years — they’ve worked for regulatory agencies, they’ve been compliance officers, they’ve been traders in the market — and they basically explain to me, “We want to detect this phenomenon in the data, something we know people do.”
And I work with, “What is the mathematical structure of that thing? What would it look like? What are the universal features of that kind of behavior?” So we work together very closely to come up with a calculation that you can implement fast.
In our first year, and this is part of working in a startup, there was a lot of experimenting as to who we were going to be and what our product was going to look like and what our team was going to be. We’ve re-geared and re-calibrated, and are much more focused.
We just moved into this space (on the 17th floor of the Chicago Board of Trade building). This is my data science lounge. When I’m hunched over my laptop and working, I need my little comfort space. You know, the mind going with the music.
If I’m hardcore debugging some code, I have some really pounding, fast music. If I’m really trying to figure something out, I’ll have something relaxed on, something slower. I have a couple of Nina Simone songs, “Moves Like Jagger” by Maroon 5, some Brandi Carlile. It kind of depends what part of the brain you’re trying to get going. My wife thinks I have terrible music taste in music.
I was born in Fort Lee, Va. I am an Army brat. Pretty much every summer of high school, I went to math camp because I was that cool. I did my bachelor’s at MIT (Massachusetts Institute of Technology) and that was in math and a minor in literature — got to keep both parts of the brain working along. I did a master’s in mechanical engineering at MIT.
College was like 42 percent women, and I lived in an all-women’s house. I would say it was after college that I started finding myself in what I would describe as more 10 percent, 5 percent (female) environments. I mean, by this point, I damn well better be comfortable in highly male environments.
I met somebody who works at Neurensic at a data science meetup in Chicago, and I was like, “Wow, you work for a company that does machine learning — machine learning and artificial intelligence are basically synonyms — and I’m so interested in that!” And she’s like, “Give me your resume.” And two weeks later, I had a job offer.
I have met some really fun, great people here. We respect each other’s personal lives. I have a 15-month-old daughter, and work-life balance is important to me.
I have a bunch of research papers I’ve written, and I love all my research papers like they are my children. But sometimes you write a beautiful research paper, and you know the impact of this paper is a certain size. Some people are going to read it. A lot of people aren’t.
When I find a pattern of manipulation inside the data that we haven’t seen before, and now we know that we’re going to catch it, the feeling that you are going to make a positive impact on the world is so much higher. I think it is really important that the financial markets be a thing that we can trust. That’s almost like a foundation of our how our economy works and how our society works, that fundamentally, these are trustworthy spaces.
As told to freelance reporter Erin Chan Ding. Stories are edited for length and clarity.