"My job starts where Excel stops!"
They are often called data geeks who are able to work magic. They juggle massive amounts of data, have excellent software development skills and have mastered at least one programming language perfectly. They are solution architects, they make complex data sets easy to understand as well as rendering them visually accessible, and they design high quality forecasting tools to support Allianz in improving its product range. Who are we talking about? Data scientists of course.
We spoke to Andreas Cardeneo, Alexander Höweler and Steffen Lotter about the diversity of the data science profession at Allianz Germany. Before joining Allianz, Andreas worked in a major German software company as a data expert; in the Advanced Analytics Department at Allianz Life he can give his capabilities free rein. Alexander and Steffen, on the other hand, are not classic text-book data scientists. Alexander didn’t graduate in data science but rather studied a classic business administration degree, albeit with a focus on data, and it was during these studies that he discovered his flair for big datasets. After gaining initial professional experience as a data scientist in IT consulting, he has been working in SMC Portfolio Management, i.e., working with small and medium-sized companies, at Allianz Germany. Steffen is an actuary at Allianz Private Health Insurance and this has increasingly brought him in touch with data analytics for some time now.
What’s the difference between an actuary and a data scientist?
Actuarial science, certainly the best-known profession in the insurance industry, and data science are often, wrongly, lumped together. Unlike a data scientist, the classic actuary mainly deals with pricing, product development and (risk) model construction and improvement. Particularly far-reaching actuarial knowledge is extremely important for this work. By contrast, data scientists are experts in data analysis and thus complement the specialist expertise of actuaries perfectly. A data scientist has to understand and be able to explain masses of data as well as transferring them to the business world and then using them to derive management strategies and solutions. We find an interface between these two fields, for example, when data science and data analytics help actuarial services to understand and anticipate customer behaviour better using analytics methods.
“First of all I have to understand the business objective and then translate it into a data science question. For example, if the initial question is “Which customers take out retirement provision products that invest in the capital markets?” then I have to develop the question “How do I classify customers into those that take out retirement provision products that invest in the capital markets and other customers?“.
– Andreas Cardeneo
What are the typical tasks of a data scientist?
There is no typical working day in the life of a data scientist. As we have already seen, the broadly based portfolio of tasks, which ranges from data analysis to presenting ideas, makes the work of a data scientist very diverse and varied. They often work together with colleagues from other specialist departments in order to collect the necessary information they need to develop the statistical models that are often considered to be the core activity of data scientists. Alex says, though, that the work is only as diverse and varied if one is pro-active. At Allianz, he adds, data science is so new that it sometimes takes a lot of persuasion before anything new can be tried.
To better describe the varied work of a data scientist, Alex uses a fish metaphor: “Data scientists work under water, programming, and surface now and again to network with others, to present current ideas and gather new ones (you always need to come up for air). Then they disappear beneath the surface again to build new models.” Thanks for the useful explanation, Alex! :-)
Alex is currently working on a task that aims to simplify the search for the right insurance cover using data sets. Let’s let Alex explain it himself: Basically, it’s very complicated for corporate customers or Allianz intermediaries to find suitable insurance cover. It means a relatively high workload for the customer and for Allianz. We want to simplify the whole process by automating many things in the proposition process for customer and intermediary using machine learning and natural language processing algorithms, and running and completing the complex calculations in the background. This means that customers can be offered the most appropriate product without even having to articulate their wishes. If the customer then says: “Great, that was simple, intuitive and just what I needed“ then I’ve done my job. The customer experience has to be simple. We take care of the complicated stuff behind the scenes. This means that customers can be offered the right insurance and, at the same time, they know they are getting an innovative product “. At the end, the challenge is to explain what we have developed so that the project partners understand the findings of the model application and know what value-added it provides.
What is special and extraordinary about being employed as a data scientist at Allianz and / or working in the field of data science?
Our interviewees emphasise the broad spectrum of topics and the diversity of the profession: the mix of persuading, innovative technology and actual implementation as well as the whole process -- from data analysis to the finished product – that one can experience.
“It’s great when you are ultimately praised for a tree, whose seeds you planted and which you watered as a sapling, and when you see it bear fruits.“ – Alexander Höweler
Steffen enthuses about the technology that Allianz Deutschland uses: “It enables us to work quickly and to reach our objectives.“ Furthermore, data scientists have the opportunity to develop themselves further, personally or in terms of their technical expertise, thanks to the abundance of training and development offered by the company. Alex is motivated on Monday mornings by the fact that he is helping Allianz to develop as a company, in lots of small steps, and to transform itself when it comes to being future-proof.
“At the end of the day it‘s about improving the bottom line of the company. That’s actually what everyone always wants; it’s nothing specifically for data scientists. The question is just where do you begin and, later, where the business processes leave their traces in the data and one has recognised that many data analyses have something in common and you can often use the same, or at least similar, methods and this has developed into a field of its own.“ – Andreas Cardeneo
Steffen enjoys the multi-faceted aspect of the profession: “New challenges are constantly emerging. Analytics methods and the underlying mathematics are also constantly developing further. There are more and more technical possibilities we can draw on. One thing’s for sure -- it’s never boring.“ – Steffen Lotter
Thanks to the different topics you work on and the different departments that you work with, your team changes all the time: you have an agile team so to speak. You may even end up working with international colleagues at other Allianz companies.
Finally we ask Steffen what he would say to young, dynamic data scientists who are interested in working for Allianz, for example, what challenges does he face every day.
“Constant new challenges require new solutions to be developed all the time. Working with big data sets at Allianz isn’t always easy because different data may be stored in different systems. Combining and validating this data isn’t always easy. But I love working with this data.“
Andreas also found it a challenge to work as a newcomer to the industry. It was a little bit like “working in a foreign country whose language you didn’t speak“, he says. But the friendly team and the great project atmosphere soon helped him to feel at ease.
“The greatest prejudice about the profession of data scientist, which I’ve always wanted to eradicate, is that you sit in the basement like a computer nerd and never speak to anyone. And that you spend 10 hours a day programming!“ – Alexander Höweler
Thanks to Andreas, Alex and Steffen for this interview and for giving us some insights into the fascinating world of the data scientist!
This article was provided by actupool Sponsoring Partner Allianz Deutschland AG. For further articles please visit the Allianz career blog.
Against the backdrop of the highly dynamic field of Data Science and Data Analytics, at its virtual Members Assembly in November 2020, the DAV decided to launch the additional qualification of Certified Actuarial Data Scientist (CADS) for members of the Association. To obtain the designation of CADS, four exams have to be passed in the subjects ADS Basic, ADS Advanced, ADS Immersion and ADS Completion. The first three subjects can be taken as specialist subjects during DAV training. Whilst the first two subjects, Basic and Advanced, require candidates to pass conventional examinations, the Immersion and Completion blocks can be passed by means of practical exams. All four modules require candidates to attend compulsory three-day preparatory courses. Among other things, the training covers data processing and data protection, developing data science applications and using state-of-the-art programming language in an actuarial context. You can find more information about the CADS qualification here (German language).