Hi I'm dgenak, and I work across machine learning, data analysis, and software development. I enjoy projects that mix statistical thinking with hands-on building, from modeling human decision-making to optimization algorithms, alongside full-stack development (Java/Spring Boot, Python/FastAPI) and basic DevOps work with Docker and Kubernetes.
What drives me most is curiosity: I like exploring new technologies and experimenting just to see how they work, which has led me into pretty different areas, from behavioral economics and NLP to Garmin watch face development.
Human-Decision-Making-Under-Uncertainty-Project: ML model predicting human choices in gambles, based on Cumulative Prospect Theory (Tversky & Kahneman) and the c13k dataset. Tech: Python, pandas, scikit-learn, LightGBM, Optuna, SHAP
Age-and-Gender-Distortion-Project: Statistical analysis of age/gender bias in GPT-2 generated text (correlation, regression, ANOVA). Tech: Python, pandas, statsmodels, scipy, seaborn, plotly
Doughnut-of-Social-and-Planetary-Boundaries-Project: Reproduces the Doughnut Economics model with custom charts, including a hand-built interactive doughnut plot. Tech: Python, pandas, matplotlib, numpy, ipympl
Capacitated-Team-Orienteering-Problem-CTOP-: Vehicle routing/delivery optimization solved with greedy construction, local search, Iterated Local Search, and Large Neighborhood Search. Tech: Python
course-match: Course comparison tool using NLP. Java (Spring Boot) backend plus a separate Python (FastAPI) NLP microservice, a clean split between app logic and the ML-heavy work.
ERASMUS_APP: Java (Maven) app with CSS styling, 35+ commits.
ERASMUS-PROJECT: Another Java (Maven) project, likely a parallel or earlier version of ERASMUS_APP.
ERASMUS-: Early prototype/demo, predating the other ERASMUS repos.
Piraeus-Bank-Project: Educational front-end mockup of a banking system (explicitly unaffiliated with the real bank), with pages for parameter selection, reporting, approvals, and FX tracking. Tech: HTML, CSS
seip_assignment_1_2026 (fork): Node.js app containerized with Docker, deployed to a local Kubernetes cluster via Minikube (ConfigMap, Secret, Deployment, Service), plus a GitHub Actions CI workflow.
DG_GarminWatchFace: Custom Garmin watch face. Tech: Monkey C
autoemulate_DG (fork): Fork of the Alan Turing Institute's AutoEmulate package (Python).
| Category | Technologies |
|---|---|
| Languages | Python, Java, JavaScript/Node.js, Monkey C, HTML/CSS |
| ML/Data Science | scikit-learn, LightGBM, Optuna, SHAP, statsmodels, pandas, numpy |
| Visualization | matplotlib, seaborn, plotly, ipympl |
| Backend | Spring Boot (Java), FastAPI (Python) |
| DevOps | Docker, Kubernetes, GitHub Actions |
| Build Tools | Maven |
