Michael
The early career path from school internship, the dual study programme in cooperation with ESG, through to permanent employment.
I first heard about ESG many years ago as a pupil at a fair at Maisach secondary school. As I love innovation as much as ESG, I quickly realised that I belonged here. Back then, I started with a school internship and then, after I finished school, I chose the dual study programme in computer science together with ESG as a partner company. It was an advantage that I was already well networked within ESG and knew exactly what I wanted to do in the future. The dual study programme alternated between theoretical training at the Munich University of Applied Sciences and practical phases at ESG's company headquarters in Fürstenfeldbruck.
It was great for me to be able to apply the theoretical knowledge from my computer science degree in practice at the same time. The collaboration was so successful that I decided to realise a project in the field of AI systems together with ESG as part of my extra-occupational Master's thesis. I find the topic of AI absolutely fascinating, because although AI systems are increasingly shaping social life and the areas of application are becoming more and more diverse, they are not infallible. They can even be deliberately disrupted by adversarial attacks. These attacks are used in image recognition systems, among others, and pose a major risk to decision-support or autonomous systems.
As part of my final thesis, I ultimately developed a demonstrator at ESG that shows the use of such an adversarial example, i.e. uses a specific application example to demonstrate how the disruption of the AI works.
It was an exciting and thoroughly successful project that gave me the feeling of having created something meaningful myself. Based on this conviction, I was happy to accept the offer of a permanent position and have been working as a Data Analyst at ESG in Fürstenfeldbruck near Munich since April 2023.
As part of the CODE Annual Conference 2023, I was allowed to represent my employer and give a presentation on the topic of Generating Background Adversarial Examples to Prevent Object Classification. This involved a method that can be used to successfully disrupt AI systems. The audience was highly interested and that way some significant conversations arose about it. It's a great feeling and I'm looking forward to my further development at ESG.
