Before joining Swisscom, you were working in academia at the EPFL in Lausanne and at HEIG-VD. What motivated you to start a career in the private sector in data analytics and AI?
During my Ph.D. at EPFL (2006-2011), I was very focused on fundamental science and was working on designing new algorithms for distributed systems and proving that they are correct and performant. I was confronted with applied science at HEIG-VD (2012-2015) and figured out how science can be useful in everyday life. Analysing massive amount of data in an efficient way and making sense of big noisy data fascinated me even more in the applied science field. I followed several online courses both in big data, machine learning and text mining and decided to continue my career in this amazing area. Here at Swisscom, we have both, a large enough set of real data to investigate, as well as stimulating problems where the goal is to make people’s life easier.
What excites you most when working with data?
The data scientist job is a multidisciplinary job. You work with different people with different backgrounds and learn a lot of things together and grow together. Also, I like the diversity that exists in data science tasks. Depending on the project that you are involved in, the domain knowledge can vary completely.
Which skills would you regard as vital in your current role and which technologies are you primarily using?
We use open source technologies as much as possible. For prototyping, either Python or Scala Spark are popular depending where data resides. In production, depending on the application, Scala Spark is mainly used for ETL tasks and batch processing, otherwise, Java and Python microservices are very common.
In my position, soft skills are as important as the technical skills, such as, ability to understand business needs, to communicate with non-technical people, to collaborate within and across teams, as well as, caring about how internal processes can be optimized, and how business needs can be translated to technical know-how.
Data science is a quickly moving field. How do you keep up to date and which online or offline sources are you using?
At Swisscom we have the tradition of organising some training courses internally, as well as organizing different community of practices (CoP). Specially in the data science CoP, we organize regular reading groups in which participants discuss the most recent academic papers or technologies and challenge how and why they can be useful in our projects. Swisscom offers 5 days of professional development per year in which you can attend a conference, workshop or training and stay up-to-date.
Furthermore, Swisscom just started a new collaboration with the EPFL Extension School and offers data training courses for data novices as well as data professionals. In the pilot project, 100 Swisscom collaborators can attend a course and the company covers the costs. Each course leads to an EPFL certificate of completion or an EPFL Certificate of Open Studies - an officially recognized academic qualification.
Which three pieces of advice would you give to aspiring data scientists?
- Talk to the domain experts in your company and try to understand their problems and needs, translate those needs to scientific problems, develop scientific solutions and analyse the equivalent business value with your business partners.
- Start with simple solutions and add complexity only if needed. When developing new solutions, think twice about operation and long-term maintenance.
- Try to find an answer to the question "why this is happening" instead of only predicting what is going to happen!
Thank you for your time!
