A Day With a Data Scientist

Data Sciences Blog Article Part 2 QAD DynaSys

We are back with Dr. Diadie Sow for the second part of our series dedicated to Data Science. In the previous post, he discussed the rise of the data science profession.  In this article, we will talk about what a Data Scientist such as Diadie is doing on a daily basis.

What qualifies a “good day at work” in the field of data science?

It is difficult to define any given day as a “good day” or a “bad day” coming from someone who is passionate about data science. However, it is important to keep in mind that data quality issues can have a huge impact on data science projects. When this happens, my colleagues and I must find the exact root of the problem to be able to understand and correct errors. Oftentimes, this can be very expensive and time-consuming. It is important to know that incorrect or inconsistent data will always lead to poor performance and/or false conclusions. When confronted with datasets with a lot of quality issues and a dataset supplier who is unable to relay accurate information, you can almost always count on there being delays and maybe even some frustrations. Conversely, properly constructed and interesting datasets make for a very enjoyable day. 

Do most days follow a routine or does every day tend to be different?

This definitely depends on where you work. If you are a consultant working for a company that has set projects in various fields such as image classification, face detection, natural language processing, etc., or as part of a very active R&D team, no there is little to no routine. Every day is different and has its own set of challenges and discoveries. Otherwise, yes, a bit of a routine is a given. As the industry average goes, the majority of data scientists tend to stay around two to three years with a company if they are not actively involved in their R&D processes.

What percentage of your time is spent researching and educating yourself about new methods and tools? 

I tend to follow the latest developments in the field via my LinkedIn contacts, on Kaggle.com, and with the publication of new packages and new books, to name a few. I also try to keep tabs on other happenings, such as natural language processing (NPL), for example, even if certain topics may be a bit removed from my current job. My evenings at home are often reserved for research and my own personal training. It is hard to give an exact percentage, but there is no denying that the field of data science moves incredibly quickly. As a researcher with a heavy mathematics background, I am interested in both the methods and the execution, specifically in the form of the math behind the latest innovations; this is very important in order to be a successful data scientist. So, all of these reasons are why training and education constitute a good percentage of my time.

How team-oriented is your work, or do you mainly work alone?

My working time is split between projects with our clients and the development of our software solution. For the latter, I collaborate heavily with the demand planning team and the manager regarding the reflection of the functionalities to be added within continuous improvement. Customer projects require less intervention from other colleagues, but there is still a need for teamwork, notably when dealing with clients’ data recovery.

In the next episode of our series dedicated to Data Science, Diadie will explain how Data Sciences help to improve Supply Chains. 

Diadie Sow
Diadie has a master degree in applied mathematics and computer sciences. He also obtained a master of science in pure mathematics, a master 2 in applied mathematics with option "Optimization, game theory and modeling in economics "at Polytechnique X and a master in mathematical modeling with option "Optimization and operational research" at the university Paris 1-Panthéon Sorbonne. After that, Diadie got a PHD in computer engineering about "Prognosis of complex systems for proactive performance management" at l'Ecole des Mines at Alès. He started to work on Machine Learning there and is now continuing his work at QAD DynaSys, where he “co-experiment” machine learning clustering and predictive algorithms with customers in order to bring, in near future, new machine learning features to our DSCP solution.