DataCareer Insights: Interview with Mara from Mobiliar

You are currently working at Mobiliar as a Data Scientist in the Data & Analytics department.
What is your role at Mobiliar and what projects are you currently working on?

I work in the Customer Analytics team with a focus on customer interactions, sales and marketing. With the help of data we try to understand the customer better, to anticipate his next step or to create Insurance products and services tailored to the individual needs of the customer.

We use Data Science in all areas of insurance: for products and services, and Sales processes, claims, underwriting, risk, pricing & reserving. In the loss department we try to use Artificial Intelligence (AI) to automate simple damage from the recording until payment is made. We use Natural Language Processing and Deep Learning, in order to automatically determine damage classification using damage description, and further information such as the location of the damage. In addition, we evaluate the use of image recognition to record damage in the event of motor vehicle damage to further simplify and automate.

How are projects usually structured? Do people in your company work in
a team and how does the composition of an analysis/data science team look like?

The majority of our work takes place in the agile organization (SAFe) in interdisciplinary Teams which are assigned to an insurance-relevant area (e.g. sales or damage). In addition to the Scrum Master and the Product Owner these teams typically consists of Data Scientists, Business Analysts, Data Engineers and software developers. This enables the teams to independently develop and automate components from start to finish. For us Data Scientists  the interface to business is very important, in order to understand requirements and to use this knowledge in data processing to be applied. But also to enable the business and to show it what is possible with Data Science and AI and thus offer a different perspective. Data science plays a major role in almost all projects. We influence decisions and trigger actions.

What excites you most about working with data?

I am very enthusiastic about the work around data. It is an extraordinary varied work, which includes a wide range of tasks and responsibilities, and involves working with different people and job profiles. I can constantly learn new things about the business and the company in general.
For me, "working with data" means constantly encountering surprises. I'm motivated by working out solutions for complex problems, to be surprised and to continuously include, adapt and optimize the solution. With the creation of facts and simple visualizations of complex relationships we are moving more and more away from intuitive, towards data-based decisions.

I am constantly confronted with new challenges and can continue our mission: we make the Mobiliar "data addicted" every day.

Which skills would you regard as vital in your current role and which technologies do you primarily use?

The skills include programming, good statistical basics, pronounced analytical thinking, but also the ability to develop pragmatic and creative solutions. In addition, a quick grasp is needed, thinking ahead and anticipating, not only interest in data, but also interested in business. Not to be neglected are communicative abilities and the mediation between business and technology. Also storytelling skills and persuasion are indispensable.

We use different technologies depending on the area. These are above all Python, R and SAS. Adapted to the problem, we use classic classification and regression methods (e.g. with the Python Library scikit-learn) but also state-of-the-art methods in Deep Learning (e.g. Tensorflow, Keras, fastText, BERT). We bring our models directly in Python also in our Microservice environment into the production (Docker, Kubernetes, Python Flask) and integrate these for use in our systems.

Data science is a rapidly evolving field. How do you stay on top of the
field and which online or offline sources do you use?

I go to conferences and meetings. We have quarterly Data Science Meetups at the Mobiliar. Online I read articles and refresh my knowledge with further training. Mobiliar is involved in various data alliances (e.g. Swiss Alliance for Data-Intensive Services) and through labs at ETH Zurich and at the University of Bern. As data scientists, of course, we benefit from this. In addition we also supervise bachelor's and master's theses, which always gives us new impulses and promotes exchange.

Which three pieces of advice would you give to aspiring data scientists?

1. To communicate the results, it is also essential to invest in storytelling.
2. Trial and error is a daily task.
3. Data is only valuable in the business context: Interest in the business and expert inputs are enormously important to understand the data and to process and use it correctly.

How do you see the development of data science and analytics over the coming years?

For me, digitization means data, a lot of data and many new possibilities. In order to create added value from this, it is imperative to have good Data Scientists. Data science will continue to become more important. I am assuming that like in  software development also in the field of data science more and more standard tools will be introduced.

What was your career path to become a Data Scientist? What steps gave you the most important learning experiences for your career?

I graduated with a master's degree in mathematics with a focus on statistics. Parallel to my studies I worked part-time in very communication-intensive jobs. I have very broad interests, also in technology. For me it is still very important to be open and curious, as well as to be ready to learn something new again.

Thank you for your time!