Mateusz Stawicki
Scholar

Mateusz Stawicki

Google Scholar ID: uiu1b0wAAAAJ
Institute of Computing Science Poznan University of Technology
AI SafetyLLMAIML
Citations & Impact
All-time
Citations
1
 
H-index
1
 
i10-index
0
 
Publications
1
 
Co-authors
1
list available
Resume (English only)
Research Experience
  • Developed an innovative research framework for quantitatively evaluating the persuasion capabilities of Large Language Models through game-based interactions and controlled experiments.
  • Led a three-person development team from initial concept to public release, developing a cross-platform mobile endless runner, Scout's Sacrifice, in Unity with C#.
  • Developed an advanced AI-powered tax assistant, TaxGPT, leveraging natural language processing to streamline the completion of Polish tax forms, enhancing accuracy and reducing processing time.
  • Optimized Generative Adversarial Network (GAN) implementation for facial image reconstruction and inpainting, demonstrating efficient deep learning architecture design.
  • Developed a Python-based image processing utility that converts raster images to ASCII art through luminance analysis, custom character mapping, and output formatting for various terminal environments.
  • Implemented a comprehensive collection of advanced algorithms including greedy optimization, dynamic programming for knapsack problems, graph theory applications, and more.
  • Created a high-performance C implementation of an INI file parser featuring memory-efficient data structures, robust error handling, and comprehensive configuration management capabilities.
  • Developed a low-level BMP image processing library with steganography capabilities, implementing bit manipulation algorithms for secure data embedding and extraction within image files.
  • Provided a multi-threaded solution to the classical synchronization problem using mutex locks, semaphores, and condition variables to ensure thread safety and prevent deadlocks in concurrent systems.
  • Collaborated on the development of a full-stack Django web application, Coffee Finder, integrating Google Maps API for geolocation services, user authentication, review systems, and responsive design patterns.
  • Customized a WordPress solution for a Polish metalworking company, featuring responsive design, multilingual support, and an integrated product catalog with custom taxonomies and post types.
  • Developed a comprehensive WordPress platform for Poznan University of Technology's AI Summer School program, featuring event management, registration systems, and dynamic content delivery for academic curricula.
  • Enhanced a popular Minecraft modification, implementing advanced AI behavior patterns, custom entity attributes, and performance optimizations for the Rebirth of the Night modpack ecosystem.
  • Refactored and extended a Minecraft modification, Better Slimes, featuring custom entity rendering, biome-specific spawn algorithms, and configurable gameplay mechanics integrated with the Rebirth of the Night modpack.
Background
  • Software Engineer & ML Researcher, crafting innovative solutions across multiple technologies.
Miscellany
  • Interests and technical areas include Machine Learning, Unity, Python, C, Django, WordPress, Java, and more.