“I’m glad that I dared to take that step.”
Determined and combative – TU alumnus Michael Fitzke in portrait
2022/12/15 by Astrid Ludwig
Michael Fitzke is a fighter. With lots of motivation and firm determination he worked for his enrolment in Mathematics at TU Darmstadt; after all, his parents had not gone to university, and he had to take the route via an Advanced Vocational Certificate of Education. “The first two semesters were tough,” the alumnus says, “but after that, I really enjoyed it.” His sense of purpose has led him to his present leadership position with the US corporation Mars. As a specialist for Machine Learning and Big Data analysis, Fitzke works on developing AI systems in veterinary medicine with the objective to enable vets to make faster and more precise diagnoses in cases of cancer, coronary diseases, or joint problems in pets.
Everything might have turned out quite differently. As an adolescent, Michael Fitzke was all fed up with education and left school after the 10th grade. He then started his vocational training at a big insurance company. There, working with numbers and fee calculations day in and day out made him regain his love of learning, especially of Mathematics, so that he began thinking about going to university. So he decided to do his Advanced Vocational Certificate of Education.
“After going to the school library and coming across a book on the Galois theory, a subdomain of algebra, I definitely wanted to study at a university,” he recounts. In Hesse, and Darmstadt, he was allowed to enrol with such a certificate. What Michael Fitzke found exciting about a university were the scientific claim and the theoretical approach. He knew it wasn’t going to be a breeze, “but I took it as a challenge. I wanted to challenge myself,” he emphasizes.
“The first two semesters were tough"
The upsides for TU Darmstadt were its good reputation and good rankings in Computer Science. Business Administration and Computer Science were the minor subjects that Michael Fitzke chose alongside his major, Mathematis. In 2008, he started his Bachelor course in Mathematics in Darmstadt.
“I was just so excited to go to university. That was just an incredible thrill for me,” he remembers in a video interview, laughing. But he quickly realised that he was missing the basics he would have encountered in an advanced Mathematics course in high school and that he would have to do a lot of compensatory work. What was more, he was not prepared for the potential obstacles waiting for him in his studies, since no one in his family had gone to university before.
“The first two semesters were tough. I had a lot of catching up to do. But I was super motivated and definitely wanted to make it.” It was a good thing for him that in his minor, Business Administration, he was able to benefit from what he had learned during his vocational training.
Well prepared for research
By the third semester, he had mastered the hard times. “Once you’ve grasped it, Mathematics gets easier,” he laughs. After that, he enjoyed his studies to the fullest. Michael Fitzke likes to remember Prof. Michael Kohler who taught probability calculus and statistics and whose lectures triggered his interest in Machine Learning.
So in his Master’s course, this TU Darmstadt alumnus specialised on focus areas like probability theory, pattern recognition, and Machine Learning. “At that time, the course was set for my later professional life,” he says. He is not only grateful for the well-founded education in Mathematics and Computer Science that he enjoyed at TU Darmstadt, but also for the good preparation for research. “Today, I’m working very close to research and I’m able to put what I’ve learned into practice every day.”
Using machine learning to help animals
His first professional stations led him as a Data Scientist to Commerz Finanz in Munich, Deloitte Consulting, and as lecturer to the University of Applied Sciences Osnabrück. Since 2018, Fitzke has been working in Verden near Bremen for “Next Generation Technologies” which belongs to the Mars corporation. The US manufacturer of such well-known brands as Milky Way, M&M’S, Ben’s Original, Whiskas, and Frolic has more than 400 locations in 80 countries. For many years, Mars has also been active in animal health; e.g. the European veterinary chain AniCura belongs to Mars.
And this is where Michael Fitzke’s expertise in terms of pattern recognition and Machine Learning comes into play. He was a project leader for “OncoPetNet”, where a system for faster and more precise cancer diagnosis in cats and dogs was developed. “Artificial Intelligence is an important tool in order to be able to make a precise diagnosis for pets at as early a stage as possible,” he says. Only few vets are specialised on visual diagnostics, such as radiology. The Mars corporation operates one of the largest digital pathology laboratories worldwide, where digital scans and tissue samples of ailing pets can be examined.
“For each scan, gigabytes of data are produced, which an AI system can analyse much faster than a human being,” says Michael Fitzke. Whether a pet suffers from cancer or not can among other things be determined by the speed of the cellular division rate. With an AI system like the one the TU Darmstadt alumnus helped develop, cell proliferation can be quantified faster, allowing a faster diagnosis of a cancerous disease.
Encouraging people from non-academic households
With his own team, Michael Fitzke continous to work on extending AI use cases in veterinary medicine. Algorithms can also help diagnose cardiovascular or joint issues in pets. He is based in Verden, but his international team are dispersed across the globe. The TU Darmstadt alumnus loves his work. “Here, I can realize many of my interests and put my knowledge from university into practice. It’s a great job which also has an ethical component, since I can help pets.”
That’s one reason why his kids, too, like daddy’s job. He would like to be a role model for other young people from non-academic backgrounds and encourage them to go for a course of study. “I’m glad that I dared to take that step,” he emphasises.
