Andreas Wygrabek works as a freelance Data Science consultant and trainer in programming and statistical analysis. He has gained more than 12 years of experience in AI, Machine Learning, and Big Data.
Before, Andreas was working at one of the first Data Science companies in Germany.
Pioneers in Data Science
I’ve been working in the field of Data Science for 12 years. My first job title was still data analyst. I started my studies in Kassel in 2003 and the combination of subjects consisting of sociology and psychology has at first glance little to do with the skillset with which I pursue my profession today as a software developer and analyst.
However, since my focus was on empirical/quantitative research, data and statistics and analysis tools were constant companions in my studies. I’ve always had a fascination for technology, software and programming languages – together with the statistical methodological know-how of my student days, this has sharpened my Data Science profile.
In 2010 an acquaintance of mine started his own business and asked me if I wanted to work in his management consultancy. The business idea was revolutionary in those days but sounds common to us nowadays: Use powerful open-source technologies to generate insights in data and replace slow, cost intensive and limited legacy systems. In retrospect, this was a great strategy, as it was precisely the time when Data Science was born in Europe. Before that, it was more common to do data analysis using proprietary software like SPSS or SAS.
You could see that the amount of data was growing very fast, but you didn’t really know how to generate insights from it.
After my time as an analytics consultant, I wanted to take a different direction and decided to do a doctorate in the field of energy economics. There we analyzed how much renewable energy should be fed into the grid to enable optimal energy use. However, I was not very fond of the research field, and luckily, my daughter was born at that time. So, I decided to give up research and start my own business as a freelancer, which I still do today.
Data Science Freelancer
As a freelance Data Science consultant, I offer three types of services in my portfolio:
- Data analysis, model development and statistical consultancy.
- Software development in the field of data analysis.
- Training and workshops on Data Science.
My courses are mainly aimed at people who wish to deepen their knowledge of analytics.
At the start, most of my first projects were training courses. A few months later, I got involved in larger Data Science development projects as a freelancer.
From PoCs to DevOps
Looking back, there was a lot of uncertainty regarding open-source software like R or Python. Especially on the management level, open-source was a bit disreputable. (Similar to Wikipedia, there was still a lot of uncertainty when using it as a professional source.)
The early phase was also characterized by companies taking their first steps in analytics and approaching data. They tried to tap into the existing large database systems to see what could be done with it. A lot of Prove of Concepts (PoC) was created. Often the business cases were not clearly defined.
This has changed in the last years. A lot of companies are no longer in this pioneering phase and are already working on and implementing product-ready solutions.
The trend shows that DevOps occupies an increasingly large part in Data Science.
In the course of this development, the requirement profile of data scientists has shifted. Especially in the early days, a Data Scientist was further away from software development. It used to be more about creating models to have the first draft. Quick-and-dirty solutions were tested to show how inherent informationcan be transformed into business cases..
Now people think more from the beginning about solutions that can be used operationally. this trend resulted in the growing need for DevOps and Software Development Skills in the Data Science-
Development of the Freelancer Market
Especially in the last years, the Data Science market was favorable for freelancers.There were many possibilities to become engaged as a developer. However, also because of Corona, an increasing slimming of the market is evident. The projects for freelancers are becoming less. This inevitably leads to the fact that I now offer my training courses online, which were previously held exclusively on or offsite.
Nevertheless, I can very well imagine continuing to work as a freelancer in the coming years. I am delighted with the development of my portfolio so far.
The Future of Data Science
I believe that we are still in our infancy. In the course of digitalization, I think that the market will continue to grow. Particularly, when I see how little digitalized, and automated many company processes still are.
As mentioned above, software development concepts are going to make up a more significant percentage of Data Science. Furthermore, I am convinced that a particular domain understanding of data itself is crucial to work efficiently and interpret the results. This makes it easier to put oneself in the problem definition and to create added value.
Create value from data
Data Science should be thought of as end-to-end.
The essence is that you manage to bring information to business processes. Data analysis must not be an end in itself. It is a critical factor that should be seen by every Data Scientist. Analysts often tend to get excellent results, but often fail to bring these results into the processes. The added value comes from the fact that a user group accepts these results as a basis for decision-making. Data Science should be thought of as end-to-end.