March 15, 2018
Today we kick off a monthly series featuring the data scientists in the Evo team. We start with Senior Data Scientist, Giuseppe Craparotta.
Where are you from?
I am from a small town in Sicily, but I have lived in Turin for many years.
Your family background?
This may have been relevant in my choice of career. Many people in my hometown work in agriculture or fishing, but my parents manage a store, so I always lived in a store (and numbers) world! I am now proud of helping lots of store managers optimize their sales.
Universities and major?
I attended the Polytechnic University of Turin and I obtained a Master’s in Mathematical Engineering. The programme aims to develop a problem-solving mentality in engineering with the use of advanced mathematical tools. Then I joined Evo, and after a year, I started a sponsored PhD in Applied Mathematics at the University of Turin. Last spring, I was a visiting student at ENSAIT, the French textile education institute.
Areas of academic interest?
In particular, I study forecasting methods, especially for the fashion industry. Also, I deal with forecast improvement by using unstructured data like images and text descriptors.
My first publication came from my work during my Master’s, and it deals with spatial prediction. Now, I am working on a number of different papers dealing with fashion forecasting and the applications of forecasting. Our work on Replenishment is at the printers right now, and papers dealing with optimal forecast models and forecast using images have been submitted.
Why did you become a data scientist?
I think that “data” is simply the way in which we label all we can observe in the world. There are facts, and data scientists aim to understand why and how these facts happen. Moreover, in recent years it has become easier and easier to find data and to process them: I find it interesting trying to understand patterns and structures that are not visible at first glance, and I decided that my job would have to deal with this.
The favourite part(s) of your job?
The best part is when I feel that our work can effectively help other companies and other people to effectively solve real problems, simplify manual work. Also, helping companies reduce waste is a very gratifying feeling.
The most challenging part(s) of your job?
I think that the most challenging part is solving clients’ problems. In particular, this involves matching the real business issue with what the machine (and science) can do; they always give us new ideas and new possibilities to extend our tools.
Your working day?
At the moment, I spend a large part of my time on training and problem-solving with the team, so every day is different! And another chunk of my time is devoted to developing research projects. And, of course, writing my thesis!
Your favourite external tech tool (and why)?
I would say that Git is my favourite tool. It is a version control system, meaning a tool allowing every scientist to work on a copy of the codes and full change history on local machines. Having a team of data scientists working on a number of tools and clients, it is crucial that all can collaborate to modify the codes in an organized, clean way!
How do you explain what you do to non-scientists?
I usually invite other people to imagine walking along the main street in their city and thinking about the stores they find. Each of them represents a brand, thousands of items, and thousands of customers: the complexity is evident right now!
Your advice for anybody who wants to become a data scientist?
As we always advise during the recruitment process, a data scientist can be focused on data cleaning, modelling, or reporting, but whatever their area of interest, a love of programming is a big help!
About the author
Martin Luxton is a writer and content strategist who specializes in explaining how technology affects business and everyday life.
Big Data and Predictive Analytics are here to stay and we have only just begun tapping into their enormous potential.