Database Reference
In-Depth Information
Karpištšenko:
You have to follow the trail through people. You usually have
some contacts you know and so you reach out to them. You learn about their
inspirations and the professionals they think are great. This process leads you
quite fast to a large set of people. Of course, in addition to going out to look
for people, you also have to promote your company at conferences. One of
the great hires we had working for us for a while, a top data scientist who was
working on top machine learning libraries, was hired through a presentation
we did at PyData and then the Kaggle competition. Getting to him took a lot
of interviews with people I met at the conferences, with people I met at the
Kaggle challenge, and a lot of Skype calls.
As you identify the right kinds of data scientists, they should meet your team
members. If the team members are all very excited, you should hire the per-
son. You must set the goal that new hires have to inspire the existing team and
have to be better in some aspect compared to everybody else. As you do this
approach, it will start to pay back, as these new people will start to bring in
new people as well. I find it inspiring to work with people who are passionate
about what they do or who have some other reason to work other than just
financial gain. Financial gain is a second-order result: if you do the right things,
everything else will follow. So I look for people who push themselves. I look
for some progress toward self-fulfillment or whatever it is that one is after.
Gutierrez:
What's a project that you and your team have worked on in the
past year?
Karpištšenko:
The most exciting project was commercially finalizing the
product that we had. It took an enormous amount effort by our very small
team to deliver this big data ocean platform. We had to make many tradeoff
decisions. We work in a field where there are petabytes of information, so we
had to look at smart ways to not overwhelm ourselves with data size. We are
in a field where there is a high diversity of data formats and data types as well.
Again we had to navigate the landscape in order not to overwhelm ourselves
with any of those. While in a sense it was a project to finalize the product,
it is also an ongoing never-ending thing because we want to continue helping
our customers.
Delivering a working commercial product in a highly complex, new industry
with very diverse and large data sets and getting it to a state where companies
trust us enough to pay for it has been greatly rewarding. It means that you've
made many right decisions and what you've created is now able to grow and
become mature enough for the ocean industry to step into a new era of an
easily accessible data-rich environment that will allow them to make better
decisions. It's great to know that I have made an impact on those businesses
to help them be more responsible with the environment around us. And from
now on, it's a new era with new challenges as we work with many more cus-
tomers in many more different ways. So the challenges will be different.
