“There is a strong focus on bringing data-based insights into production.”

Working at ZEISS

Data Scientist Kaveh Pouran Yousef gives insights on his daily work at ZEISS Digital Innovation Partners.

Hi Kaveh, what do you like about working at ZEISS?

We build digital products and services based on the ZEISS DNA. This means combining a long tradition of creating high quality high precision products and solutions with an agile way of software development and a cloud-first approach. This also means having the best of both worlds. So I have a privilege to work with and learn from great hardware engineers as well as frontend developers, data scientists and business analysts.

Can you tell us more about your professional background and your role at ZEISS today?

I did a PhD and a postdoc in Bioinformatics, where I studied the dynamics associated with the spread of diseases. At ZEISS, I work as a data scientist, building analytic models for data originating from sources as diverse as IoT streams from connected devices or service data from classic SAP databases. Identifying patterns in complex systems and developing prediction models remains the main part of my job at ZEISS. What has changed is a stronger focus on bringing data-based insights into production.

 

 

Can you give us more insights on your work as a data scientist? Which tools are you using regularly?

I work with Apache Spark using the Databricks platform on Azure. Spark enables writing programs in Scala or Python just as if you were working on a single computer. In the background, the cluster manager distributes the job to multiples computing nodes for you. So whenever your data load becomes too large, you just need to increase the number of nodes in the cluster without changing your programming logic. This scalability really simplifies my life as a data scientist so I can pay attention to more interesting things, like the quality of our prediction models.

What are some of the key competencies a data scientist needs for his everyday work?

A crucial part of building data-based products using machine learning is communication. I need to discuss what I do and make it understandable to colleagues from various backgrounds. No matter how good a prediction model is, it yields no added value if nobody understands it. I also need to build sound understanding of the domain where my data comes from, which requires an intensive exchange with domain experts.

How would you describe a person who fits perfectly into the Digital Innovation Partners team?

We are a very diverse team, but two common traits that I see are curiosity and customer-oriented thinking. You need the former to build bridges between disciplines and mindsets, and the latter to incorporate all that innovation into awesome products. If you are interested in digital products, agile way of working, data-driven thinking, and curious what it all has to do with measuring machines or medical robotics – then you are a great fit.

One more question: Do you have a favorite ZEISS product?

This is a tough one because I have to choose here. There are some measuring machines that I could imagine having at home in my living room… Seriously, I think ZEISS O-INSPECT has a great combination of tactile measurement and optical scanning. In addition, due to my previous research work, I have an affinity for microscopy. Seeing an amazing bacterial biofilm image on one of the ZEISS electron microscopes will completely change your perception of the concepts of size and distance.

Tags: , ,

Leave a Reply

Your email address will not be published. Required fields are marked *