Projects
Three contributions to the GameMakers Toolkit GameJam
2019, 2020, 2021
My first three bigger development projects were created for the GMTK GameJam together with a team of friends.
- Developed three unique games in three 48-hour timeframes
- Utilized various game development technologies and frameworks like Unity
- Improved rapid prototyping, problem-solving skills and teamwork
Disclaimer: The image shows a standard neo4j graph, the original project was developed for internal use and is therefore not public
Network Graph: Cooperation with a bank for a group university project
2022-2023
A group university project aimed at visualizing dependencies and usages of the clients internal software products.
- Created an interactive graph to visualize software dependencies
- Improved upon an existing proof-of-concept implementation
- Utilized React Frontend, Spring Boot REST API, and Neo4j graph database
Predicting National Team Football Match Outcomes
2024
Developed a multistage training approach to predict national football matches using deep neural networks.
- Trained an embedder and classifier model on national league games
- Fine-tuned the model on sparse national team data
- Achieved accuracy around the betting odds accuracy for UEFA Euro 2024
Disclaimer: AI generated image, official logo pending
ALPACA: Automatic Library and Program API Change Analysis
2023 - Ongoing
Developed a C++ tool to identify changes between versions of large C/C++ code bases.
- Uses Clang Compiler-Frontend to extract information from the code
- Implements various static and dynamic matching methods
- Potential use in change documentation and automatic refactoring
- Version 1.0 evaluated with an average precision of 89.9% and a recall of 92.3% on among other libraries OpenMPI
Personal Deep Learning Face Recognition Sandbox
2024-Ongoing
A personal project to learn about deep learning in face recognition and computer vision in general.
- Implemented a pipeline of face detection and face recognition on live video
- Using pre-trained face detection models and a self-trained face recognition model
- The Goal was to learn about the basics of deep learning in face recognition in particular and computer vision in general