Iggy's head of data science Anne Cocos recently sat down to discuss Iggy and location ML on the ML Ops Community podcast.
Tune in to hear Anne discuss:
We hope you enjoy it! Reach out on Twitter or email w comments or questions.
I thought I wanted to be an engineer for a Formula One team (kind of ridiculous in retrospect), so at the time I majored in mechanical engineering. Long story short, I was programming every night instead of doing my mechanical engineering homework, so I eventually decided to leave to find a job building software.
My background is mixed. I was an English major originally, and pivoted to environmental science and biology later in college. In graduate school, I got interested in quantitative / computational ecology and testing ecological theory using historical observational data, experiments, and computer simulations. I worked as an applied ecologist at Northern Arizona University for a couple of years, before co-founding a company called Conservation Science Partners in 2012. We had a tremendous amount of success over the last 10 years, but I was feeling ready for a change in focus.
As I was finishing my PhD, I did Insight Data Science and got hired at Airbnb where I hoped to use geospatial data to work on host growth or something. I was working on the host team as a Data Scientist and I realized that we didn't have any concept of the “context” of our listings… we knew a lot about the listing and its specific characteristics but we didn't really have a way to understand what was nearby. That matters because if you know what’s nearby you can market listings differently (which is sort of what Airbnb is (finally) doing w Categories), you can price more accurately, and you can drastically improve search. That became really compelling to me. I tried to work on an internal project there, but it didn't get a lot of support. So eventually, I decided to leave Airbnb and start Iggy to solve this challenge.