Database Reference
In-Depth Information
Perlich exemplifies the data scientist who makes her own rules and succeeds in
multiple areas of business, marketing, computing, and academics. Her multifarious
talents are critical for her unique role in Dstillery as a scientist, mentor, and ambas-
sador for the company. Perlich's enthusiasm for learning, sharing, and mentoring both
at work and in her academic role at NYU invigorates her interview.
Sebastian Gutierrez:
Tell me about how you got started with working
with data.
Claudia Perlich:
I started out in computer science. I was sort of a math geek,
but I was more interested in applications rather than just theory. Through
an accident, I ended up in a university exchange program and came to The
University of Colorado at Boulder for a year. While selecting courses, I found
a course on artificial neural networks taught by a German professor and, on
the spur of a moment, I decided, “Yeah, why not?”
The German professor was Andreas Weigend. He later became chief scientist
of Amazon. He is one of my mentors and a great friend.
Gutierrez:
What happened after the exchange program?
Perlich:
I went back to my university really excited about data. So I dove into
it and really started working more with data. I finished my degree in Germany
and started looking for a PhD program in computer science. Meanwhile,
Andreas had gotten a job at NYU Stern Business School in the Information
Systems Department. When I asked him for recommendation letters for
those computer science programs, he suggested I apply to NYU Stern. I told
him, “Look, I don't know anything about business.” He said, “Well, neither do
I.” Through that connection, I ended up at NYU Stern where I obtained a PhD
in Information Systems, working with Foster Provost on machine learning and
data mining. My dissertation was on predictive modeling.
Gutierrez:
Did you have any exposure to the industry side of research?
Perlich:
I did a summer internship at IBM's Watson Research Center in one
of the summers during my PhD. I really enjoyed the atmosphere and the group
there, so as I contemplated my life after finishing the PhD, the decision was to
either go into academia or to go to IBM. In the end, I decided to go to IBM
and ended up working there for six years in the predictive modeling group. I
really liked the diversity of projects, because they have all kinds of internal and
external consulting projects coming in. All these projects centered around
data and building data-driven solutions. We also had time to publish and par-
ticipate in data mining competitions, so it was a great atmosphere for me.
Gutierrez:
What drove you to leave IBM for a startup?
