Database Reference
In-Depth Information
Gutierrez:
What do you consider to be the most helpful domain-specific
knowledge at PlaceIQ?
Lenaghan:
The domain here at PlaceIQ is people. What are people doing?
Who are these people? What kinds of heuristics can we layer over our data,
either intuitively or anthropologically, for it to be true? That gives us tremen-
dous mileage.
We do have a lot of analysts who short-circuit some of the algorithmic work
we do just because they know that something may or may not be true just
based upon human experience.
Gutierrez:
You keep in mind that it is real people behind the massive amounts
of data.
Lenaghan:
Correct—we want to build an analytics platform rather than give
our customers black-box answers. We really want to be used and viewed as
an augmented intelligence service for analysts. These are people who are run-
ning campaigns. These are people in the consumer insights business.
Gutierrez:
There is an idea of the data exhaust. In the operation of ingest-
ing and analyzing your data, you are also generating data that could be useful
for other people. How do you think about this secondary data? Do you think
about monetizing, giving it back to the community, or a combination?
Lenaghan:
That is a very good question. Right now, we are not doing that.
We are laser-focused on consumer insights and especially on mobile advertis-
ing. That said, I think the long-term vision of the company is the platform. It
is the platform we can license to other people. It is a platform with potential
APIs to give this contextualized information back to the community. I think
that is really the direction of the company. It is not what we are going to be
doing in the next bit, but over the next few years.
Gutierrez:
Speaking of communities, you mentioned earlier that you are
using R, Python, and Java, which are tools built by open source software com-
munities. What tools do you use and how has that changed in your career?
Lenaghan:
When I was working in trading, I worked mainly in C++ and Perl.
It makes me feel very old when I say that. Now we hire young engineers, and
they have never used C++ or Perl, and that sounds crazy to me.
Then moving into this world, I do most of my work in Python. The number
of very useful libraries and frameworks in Python seems to be growing every
day. Another benefit of Python is that you can also write prototypes very,
very quickly. Since performance is not a super big issue from the perspective
of building prototypes, I always go to Python. I would say for up until about
the beginning of the summer of 2013, I was writing a lot in Java. That was also
when I was writing a lot of the back-end code for the data science group. But
now it is all Python.
