What does a data scientist do?

What exactly does your job as a data scientist entail?

You can create a projection of how many calls your service desk will receive on a given day, so that the service desk can be staffed as efficiently as possible for that shift, with the objective of responding to customers quickly and not letting employees wait too long. You can then measure how successful you were and learn from any issues when scheduling staff for the next shift.

And what is biggest challenge?

Customers expect things to be handled quickly. If I create a Spotify account, I can use it right away, and if I order a book, it will be delivered tomorrow. As a bank, you also want to provide that same experience; however there are more risks involved and more legal regulations you have to take into account. How can we ensure that the process is as smooth as possible for the customer, while efficiently mitigating the risks and complying with the regulations?

Can you take us through an average workday?

There’s no such thing. Yesterday for instance, I worked on an analysis to assess how heavy the workload is for the team that handles administrative tasks for new clients. How much work does it take to bring in different types of new clients? And, how can we take that information into account on the marketing side? I then discussed the analysis with co-workers. The job is extremely varied. You must be able to communicate well, so that you can get to the root of a problem, then you have to brainstorm and put the puzzle pieces together to come up with a good solution. After that, you have to communicate well again to proceed to the appropriate next step.

What is your best tip for future data scientists?

Be critical of what you spend most of your energy on. Make sure you know ahead of time what can be done with the results, and spend most of your time on those areas where you expect the most added value. In some cases, a simple calculation is enough, even if it would be more fun to build a complex model.