Welcome to my portfolio! I’m Martin, a final-year Computer Science and Engineering student at Ss. Cyril and Methodius University. My work bridges the fields of AI, ML and DS. I enjoy solving open-ended challenges and building systems that combine intelligence, automation, and creativity.
Explored the use of transformer models in internal tools to improve automation and efficiency in telecom operations.
- Experimented with Hugging Face transformers and prompting strategies
- Focused on domain-specific adaptation and chatbot prototype integration
Developing a supervised learning model to predict football match outcomes based on historical results and team performance statistics.
- Web scraping, data preprocessing, feature extraction, and model selection
- Evaluating accuracy and feature importance for real-world match prediction
Building a project based on Graph Neural Networks to explore structural data relationships.
- Applying GNN architectures on graph-structured datasets
- Focusing on representation learning and node classification tasks
Worked on the development of scalable data workflows and monitoring systems to support real-time insights for telecom infrastructure.
- Involved in building data pipelines and dashboards using open-source tools
- Focused on automation, reliability, and cross-database integration
- Created visualization tools to monitor infrastructure performance
- Integrated secure polling and automated alerting mechanisms for internal use
BSc in Computer Science and Engineering
Ss. Cyril and Methodius University – FINKI
Expected Graduation: 2026
Focus Areas: Machine Learning, Reinforcement Learning, AI, NLP, Software Engineering, CNNs, GNNs
- Languages: Python, SQL
- Frameworks/Libraries: Hugging Face, PyTorch, scikit-learn, Pandas
- Tools: Docker, Apache Airflow, Grafana, Git
- Concepts: MLOps, AutoML, Prompt Engineering, Data Pipelines
🚧 Please note: Some projects are not yet publicly available on GitHub due to internal or research-related constraints. Descriptions provided here are general and meant to highlight my role and contributions without revealing confidential details.
- Email: martin.stamenov03@gmail.com
- LinkedIn: linkedin.com/in/martin-stamenov-250ab5200
Feel free to connect or reach out – I’m always open to learning, collaborating, or discussing interesting AI ideas!