I'm Stefan, a computer scientist crafting creative, efficient, and innovative solutions. I stand for:
I love to implement modern solutions and explore ways we can solve real-world problems with innovative approaches.
I believe there is tremendous potential in bringing machine learning to your business and everyday life.
Machine learning (often referred to as Artificial Intelligence (AI) or Künstliche Intelligenz (KI) in German) has
many use cases. Currently, it is widely recognized for its applications in large language models (LLMs) like ChatGPT.
This makes sense because its performance in knowledge work is amazing. Personally, I use it more often than Google,
as the results are, on average, better than reading a random blog post about a topic.
However, besides LLMs, we should not forget the other great applications of machine learning. One of these is pattern recognition and prediction. Give it a try below :)
If you are wondering why it just won't correctly predict your perfect 1 (I did, at least), that’s because of a fundamental rule of machine learning: It’s all about the data. In the MNIST dataset, the number 1 is often written as a |, so our poor model never really had a chance to correctly recognize it...
The UI and main functionalities of Neutrino Review:
As a Full Stack Developer at CERN, I worked with an international team and in collaboration with the WHO, contributing independently to key areas of the CAiMIRA systematic review project which reduces the workload in systematic reviews by about 68% which enables a 50% decrease in the time needed to complete the screening process (Source). Here are the main aspects of my work:
My database design which focuses on ensuring scalability and flexibility:
This role allowed me to collaborate with an international team while working independently on key project components. Through this process I was able to hone following skills:
Special thanks to Elias Sandner, Igor Jakovljević and Mario Negovetić for their guidance and support, which greatly enhanced my understanding and learning.
At Itell.Solutions, I worked in different roles across three time periods: July 2019, February 2020, and August-September 2020. Each phase contributed significantly to my professional growth:
This role allowed me to develop both soft skills and technical expertise, which I continue to leverage in my professional journey. Key skills and tools I honed include:
I would like to express my gratitude to Dominik Hütter and David Prott who teached me a lot about business and sales as well as a huge thank you to my mentor, who helped me become a much better developer.
Beginnings
Unfortunately, I never came into contact with coding from a young age. None of the people I knew were developers. As a teenager, I was afraid that I could destroy my computer by doing something wrong, so I never tried actual coding, although I loved doing things on my computer. I attended a Handelsakademie (HAK) and chose the IT focus there. But here as well, we never did more than a double for loop in VBA. So my first contact with "real" coding was at university.
This resulted in this hilarious tic-tac-toe game, where I wrote 864 lines of code ... xD
Intermediate
After gaining a couple of years of experience through university projects and internships, I reached a point where I needed to address a personal challenge. I often played computer games late into the night, which negatively impacted my academic progress. However, I couldn't find any tools that met my needs—or at least, none that I was comfortable granting the necessary privileges to. As a result, I decided to create my own solution: auto_close_programs. This was an exhilarating experience because it allowed me to solve a real-world problem in a straightforward way. It taught me that a useful program doesn’t always have to be overly complex.
After that, I explored various branches of computer science, diving into areas like game development, information security/ethical hacking, and networking. For instance, I experimented with setting up a Minecraft server on a Raspberry Pi and explored other use cases for the device. These experiences allowed me to broaden my skill set and gain hands-on exposure to different fields.
One of my game development projects: A 3D RPG created with C# and Unity.
Future
After exploring many different fields, I’ve discovered that machine learning and full-stack engineering are where I truly belong. These areas resonate with me because they offer the potential to solve a wide range of real-world problems. I’m particularly excited about the possibilities we can unlock by combining machine learning with robust engineering practices to create innovative and impactful solutions. I look forward to contributing to this ever-evolving field and seeing what we can achieve through the power of machine learning and code.
I was part of the TU Graz Data Team for three terms.
In the summer of 2022, I collaborated with Hussain Hussain on the WSDM2022-Challenge. During this time, I gained valuable insights into graph-based methods from Hussain. However, I wasn’t able to contribute as much as I would have liked, as most of my time was dedicated to other subjects.
In the winter of 2024, I participated in two challenges. The first was the Xeek - Two Birds, One Neural Network challenge, where we achieved first place. I must note that Thomas Wedenig carried out the majority of the heavy lifting in this project. The second was the Optiver - Trading at the Close Kaggle challenge, where our team secured a silver medal (top 5%) using LightGBM.
In the summer of 2024, I worked on the Home Credit - Credit Risk Model Stability Kaggle challenge, where we ranked 1443 out of 3856 participants. This challenge was particularly difficult due to frequent changes in the evaluation criteria and the fact that the system was hackable, which ultimately led to a loss of motivation within the team.
Introduction
This project is meant to showcase my development process. One key lesson I’ve learned is that strong fundamentals are essential—after all, you can’t build a house on sand, right? With a solid foundation, building high-quality, scalable, flexible, and robust applications becomes more straightforward, albeit not necessarily easy. The goal is always to ensure the application can grow and adapt in the future.
My Process
Here’s how I approached this project:
My initial design:
Reflection and Learning
Through feedback from friends and self-reflection, I’ve gained a clearer understanding of where I need to improve. For example, I realized that frontend development and writing required much more effort than I initially anticipated. This realization highlighted the need to strengthen my fundamentals in CSS, JavaScript, and TypeScript. While my expertise lies in backend development and machine learning, continuous reflection and learning are critical, no matter how experienced we are.
This project reflects not only my technical skills but also my ability to plan, iterate, and adapt. It’s a testament to my belief in building strong foundations and continuously improving through feedback and learning. While there’s always room for growth, this process has allowed me to sharpen my expertise in multiple areas and create something I'm proud of.
Programming Languages
Web Dev and CI/CD
Computer Science
Non-Technical
Health is the most important thing we can have - most people would agree. Unfortunately, there’s so much misinformation out there, and many people find topics like diet confusing (myself included). I’ve tried many diets - keto, carnivore, Mediterranean, vegan, and more. What’s interesting is that we’ve known the rough steps to a healthy diet since the 1980s. Of course, if you have allergies or specific illnesses, these findings might not apply to you. But for 90% of the population, they do. According to Dr. Gardner, the 7 key principles are:
Interestingly, recent research shows that even small amounts of alcohol can be harmful. What’s great about these 7 principles is how straightforward they make it to figure out if something is healthy or not. You can even break it down into numbers: sugar should make up no more than 5% of your daily calories, saturated fats no more than 10%, and salt should stay under 2.3 grams. For someone eating 2,500 calories a day (the average male), that’s about 32g of sugar and 27g of saturated fat. Let’s put this into perspective: a lot of people think being vegan or vegetarian automatically means eating healthy. But if you’re having a vegan cake, it could still be loaded with sugar and saturated fats. The same goes for meat substitutes, which often have way too much salt. So, always check those 7 simple guidelines, aim to stick to them as much as possible, and you’ll be on the right track.
Personally, I’ve found that a plant-based whole food diet with 1–2 servings of meat or fish per week (basically a Mediterranean diet with a plant-focus) works best for me. It naturally fits within the 7 principles while still being flexible enough to enjoy meals out with friends.
Beyond diet, sleep and exercise are crucial for a healthy life. In a nutshell, aim for 7–9 hours of high-quality sleep, 2–3 hours of weightlifting, and 2–3 hours of moderate cardio per week for optimal health. I’d also add ergonomics (workspace and sleep setup) and stress management (e.g., meditation, yoga nidra, progressive muscle relaxation).
Great sources for the above:
The amazing thing about science is that it will work for most people. This way, we can make great progress with a fraction of the time investment compared to following random influencers (who might even give us harmful advice) and trying everything to see what works for us.