Database Reference
In-Depth Information
Sebastian Gutierrez:
Tell me about where you work.
Amy Heineike:
I work for a startup called Quid that has built an intelligence
platform. The idea is that we're going out and getting lots of big, complex,
unstructured, and interesting data sets, and then we're building analytics and
visualization on top of them so that we can use the whole platform to enable
people to connect ideas and events, tell stories, and discover what's happen-
ing in the world. I am a bit of a data geek, so I find it very exciting to get to
explore these big data sets that you don't normally get to look at analytically.
For instance, you normally get to read one news article at a time. Now we get
to actually sit and think about what would be possible if we could read a ton
of news articles at the same time. It's data science heaven.
Gutierrez:
What's been your career progression and how did you come to
your current position at Quid?
Heineike:
I started out in mathematics, where I studied a broad range of
different mathematical topics. Through my studies, I became fascinated by
graph theory, by nonlinear dynamics, and by the question of what these two
subjects tell us about how human and social systems work. So when I gradu-
ated, my first jobs were in an economic consultancy in London. In these roles
we were tasked with thinking about questions like: How do cities evolve and
grow? What happens if you're building a train line across a city? Does that
affect the city in a discernable way and what is that worth in money? We got
to ask some really fascinating questions.
I was very fortunate in that one of the big projects we worked on was Crossrail,
which is now actually getting built. I was able to swoop in after 30 years of an
ongoing debate in Britain about whether it was worth building this new train
line, and be there for the day when they actually decided they were going to
go ahead. In those roles I was doing a great deal of mathematical modeling and
a lot of data analysis, as well as also making (and then explaining) the case for
what we thought was going to happen.
What struck me at the time was how constrained we were by the data that
we had. I'd come in fascinated by the mathematical models we were starting
to build about how cities might evolve and what kinds of nonlinear interac-
tions might happen. But we couldn't actually really get into that depth because,
at the time, we were using mainly survey data. This means that you'd liter-
ally send out a bunch of people to stand on street corners and count the
cars going past on a particular day. Or we'd get survey data from the census
where people would state which locations they normally commute between.
So the data was very limited in the detail and resolution of what you could
do. However, it was very clear at the same time that data could be generated
at scale in unusual places that would make the kind of analysis we were doing
far, far better and more interesting. And so when I left that, I started thinking
about how I could closer to more data.
