Stefan Thoss: Part Time Mercedes-Benz Innovator, Part Time Student
By Giselle Diaz
Stefan Thoss has a lot to celebrate this spring. Four years ago, Stefan had an internship at Mercedes-Benz Research & Development North America, but dreamed of pursuing a Master’s Degree. After deliberating whether he should return to Germany for his Master’s or stay in the Bay Area to focus on his career, Stefan chose the best of both worlds. Two years later, Stefan has managed to work through the part-time Master of Engineering degree in Decision Analytics at UC Berkeley while also growing his career from intern to Senior Data Engineer at Mercedes-Benz.
Jumping Back Into Academia
“In the beginning it was quite hard,” recalls Stefan. “Going back to school was a lot more intense than I thought it was going to be. I was working 40–60 hours a week, and working another 20 hours on school.” However, Stefan ultimately felt that working for two years in industry helped him manage this new intense schedule.
“It’s really all about time management in the end. Working a full time job teaches you how to finish something the way it is expected. It makes you ask, “What’s essential here?’”
In some cases, however, taking a few years to work in the industry gave Stefan and the other part-time students an advantage — especially when it comes to business. “Business skills are one of the most important things and it’s one of the areas where I felt we actually had an advantage over full-time students,” Stefan says, “We present to management and colleagues all the time.” By comparison, classroom pitches were a breeze.
Driving Innovation at Mercedes-Benz
So what does Stefan do when he’s not studying hard at Berkeley? Working hard at Mercedes-Benz. As a Senior Data Engineer, Stefan is tasked with exploring new innovations for making cars a more intuitive and luxurious experience. In this role, he leverages data analytics and data engineering to learn more about drivers and how to best meet their needs.
“How customers use the vehicle is a hot topic in the industry,” says Stefan. “We learn how they drive and use that information to improve the product. For example, if we detect that the vehicle is acting abnormally, we can step in and provide maintenance. Basically, it’s centered around trying to improve the whole user experience of the vehicle.”
With the automotive industry being a rather slow moving industry, the Silicon Valley office of Mercedes-Benz is the best location to bring in fresh technology. “It’s not enough to build great vehicles. Customers want connected car features, smart phone apps, driving statistics, recommendations, and all the features they’re used to on smartphone apps. Every company is collecting data now but that’s not the hard part; the hard part is making sense of the data.”
There are many questions regarding data from cars that need to be addressed: How is data collected and transferred from vehicles? How can the data be stored and made accessible for analysis? How can the analysis of hundreds of thousands of vehicles be done efficiently? How can we learn about the customer while not invading the driver’s privacy?
“We want to understand how our vehicles are used in order to make better decisions about future car design. Another engineer in the company might have a question about how people use a certain feature in order to improve it.” This is where the vehicle data comes in. “We analyze the behavior of all customers in an anonymized way. When we discover that a feature is never used, it might be time to remove it. Or when we see that people have to use complicated menus to access popular functions, it might be time to rework the interface.”
“The fun part about the job is that I get to play around with new technology and programming frameworks such as the UC Berkeley-developed data processing framework, Apache Spark. I was the first one to introduce Apache Spark to the office after I was introduced to it at a meetup. I played around with it and now we’re using it everyday to process vehicle data.”
Sometimes homework assignments and full-time job duties are surprisingly similar. “In one class, we had to gather data from different sources, combine them, analyze them, and present the results in an easily understandable way. That’s basically what I do at my job, just on a bigger scale.”
“I was already working with colleagues that have statistics degrees and are experts at machine learning before my degree. But thanks to the machine learning courses I took as part of my degree, I can finally talk with them about kernel regressions and know what they’re talking about. Often, what I learned at school was not directly applicable to my work but helped me understand the bigger picture and other aspects of a problem.”
A broader understanding of data analytics, has provided Stefan with the base to grow in his career, but he continues to be motivated by his passion for cars and new technology. As for his future? “Autonomous vehicles are such an interesting field, and I look forward to getting more involved with data research in that sector. ”