Leonidas Devetzidis

Played a key role in developing and approving a physical device for red team engagements. Configured and hardened Kali Linux to ensure security. Deployed and managed an OpenVPN server on AWS. Practiced penetration testing on infrastructure and web apps via Hack The Box, emphasizing red and blue team strategies. Developed and documented red team infrastructure. Designed detailed PowerPoint presentations to simplify and convey complex security concepts effectively.

Title: Image Classification via Deep Learning by Utilizing Knowledge Distillation and Evaluating Explainable Artificial Intelligence Methods Supervisors: Prof. Gemma Roig and Dr. Gereziher Adhane Fields: Deep Learning, Model Compression, Explainable Artificial Intelligence (XAI) Language: PyTorch/Python Mark: TBD

Improved neural networks and survival analysis models in customer projects using Python/PyTorch. Calculated regions with the largest throughput of people to visualize optimal promotion locations using QGIS.

Fixed network and software vulnerabilities identified in penetration tests using PowerShell. Configured Windows services using Desired State Configuration with PowerShell. Automated processes using VBA and PowerShell. Analyzed network packets withWireshark. Monitored networks and configured REST-API sensors in PRTG.

Analyzed, produced and contributed to risk-related topics, including liquidity, Value-at-Risk, and leverage reporting. Improved existing reporting solutions using VBA and Python. Qualitätssicherung in der Marktrisikomessung (Derivateverordnung) und Liquiditätsrisikomessung. Prozessautomatisierung mit VBA und Python.

Title: Dendritic spine detection with Convolutional Neural Networks Supervisors: Prof. Matthias Kaschube and Dr. Jens-Bastian Eppler Fields: Machine Learning, Image Processing, Computational Neuroscience Language: Tensorflow/Python Mark: 1.1

Rollout and software migration.

Current Mark: 2,4

Mark: 2,5