Database Reference
In-Depth Information
Sebastian Gutierrez: Tell me about your journey to becoming a data
scientist at PlaceIQ.
Jonathan Lenaghan: Prior to joining PlaceIQ as a data scientist in March
of 2012, I worked in the financial services industry doing algorithmic trad-
ing. Before that, I worked for eight years in academic physics. So I've always
worked with a great deal of data—although, compared to my algorithmic
trading work, my physics work was a bit more weighted toward the analytical
than the computational.
Gutierrez: Algorithmic trading sounds like an interesting job with interesting
data sets. What drove the transition to PlaceIQ?
Lenaghan: I liked the style of work I was doing in the financial services
industry, solving quantitative problems, but it began to feel like I was solving
the same problem year after year with slight variations. After five years, I was
ready for a new challenge.
The data startup community in New York was growing very rapidly, and
I started going to the New York tech meetups. At a meetup in February
of 2012, I found myself sitting next to a VC who told me about the sort of
companies he was investing in. I told him about the work I was doing with
data in algorithmic trading. He told me that a couple of his companies needed
a data scientist and suggested I speak with them. The next day I spoke with
PlaceIQ's CTO, Steve Milton, and then a couple of days later with the CEO,
Duncan McCall. A week and a half later I started work as PlaceIQ's head of
Data Science.
Gutierrez: Wow, that was a pretty fast transition!
Lenaghan: Two weeks.
Gutierrez: Something must have really excited you about PlaceIQ. What
was it?
Lenaghan: The size and variety of the data sets. PlaceIQ is a location intelli-
gence platform, and so we ingest all kinds of data—social data, GIS [Geographic
Information System] data, POI and AOI [Point/Area of Interest] data, and
some consumer behavior data
Gutierrez: What do you do with these data sets?
Lenaghan: We've divided up the world into 100-meter-by-100-meter tiles.
We ingest all this data, cleanse it, normalize it, and project it onto our
100-meter-square tiles integrated with what we call our “PlaceIQ time
periods.” Using this unified data layer, we then apply machine learning to
contextualize locations and movement data.
 
Search WWH ::




Custom Search