Greetings! I am a researcher in the area of Artificial intelligence, with a focus on NLP, generative AI, and deep learning, working at the Distributed and Operating Systems group at the Technical University Berlin. Particularly, my research centres on developing methods for analysis of heterogeneous data types from complex IT systems (e.g., data centres), with technical objectives related to enhancing their reliability. I am also interested in meta-learning, and multi-label classification. As an applied scientist, I closely collaborate with Huawei Munich Research Center's Ultra-scale AIOps laboratory, allowing me to work on industrial projects for more than 4 years, and conduct my research using production data.

Short Biography

I hold a PhD in Computer Science from Technical University in Berlin, obtained under the supervision of Prof. Dr Odej Kao . Currently, I lead the Huawei-TUB Innovation Lab, a joint research project dedicated to enhancing the reliability of complex IT systems through AI. Previously, I was a research associate at the Department of Knowledge Technologies at Jozef Stefan Institute in Slovenia, under the supervision of Dr Dragi Kocev and Prof. Dr Saso Dzeroski, where I completed my master's studies in computer science. During my role as a research assistant, I made significant contributions to various projects, such as the comprehensive study on multi-label classification, the largest to date, and ultimately led to a tutorial at ECML PKDD 2021 on the topic of [FAIR MLC 2021].

Experience

One of the most frequent data types I have worked with is event-type data represent as texts, with some additional strucutre (e.g., context dependacy from source code, or operational context given from current system workload, or configuration files and other things like that). I have also extensive experience with multi-label classification and meta learning. Some of the code for my methods is available on GitHub. If you are interested in a particular method, please contact me.

Domain of Expertise

I enjoy learning, reading and obtaining knowledge in many different areas in artificial intelligence, machine learning and data science. I find that is important to pay attention to both, existing, but as well as prior work in the areas. Alongside my work, I have been implementing methods from the literature, from linear regression, through optimizers, up until fency GPT-3 style languge models, or vision Transformers alike. I have been using many of these methods through my work, or for sharpening my knowledge and making them publicly available at [Machine Learning Papers with Code]. This justifies the following list of experiences:
  • Generative AI, with main focus on text data
  • Deep Learning
  • Machine Learning
  • Statistics
  • Optimization (LP, QP)
  • A/B Testing and Experiment Desgin
  • Time Series Analysis
  • Multi-label Classification
  • Recommender Systems
  • Time series analysis
  • Causal Learning (basics)

Technical Experience and Skills

The research and developing aspects brought me to the usage and extensive command of the following tools:
  • Python
  • PyTorch
  • Python Analtics Stack
  • Data Structures and Algorithms
  • Optuna
  • MLFlow
  • CI/CD
  • Docker
  • Git, Github Actions
  • Relational Databases
  • SQL
  • Hugging Face Hub

Projects

The following is a list of projects I have been working on. I have organized them by the major topics I have studied so far, so within each of the repositoris you can find a bunch of subprojects, each addressing particular task or task property.

Service to the Scientific Community

Organization of Conferences and Workshops

Paper Reviewing