AI Student · Builder · Bridging technology and people
B.Sc. Artificial Intelligence @ Deggendorf Institute of Technology
I work at the intersection of AI and software engineering — but what drives me just as much is the human side of things. I've spent years collaborating across cultures, languages, and disciplines, and I genuinely think that shapes how I build and problem-solve as much as any technical skill does.
| Project | Stack | Description |
|---|---|---|
| 🩺 CheXVision | PyTorch, CNNs, DenseNet-121, Python | Deep learning pipeline for multi-label classification of 14 pathologies from 112,000+ chest X-ray images (NIH ChestX-ray14). Compared a custom CNN architecture with transfer learning and evaluated models using AUC-ROC and F1-score. |
| 🚆 Deutsche Bahn Train Delay Prediction | scikit-learn, pandas, Python | Supervised regression study using 2M+ Deutsche Bahn records. Compared Linear Regression, KNN, and Random Forest with domain-specific feature engineering and forward stepwise selection. Report |
| ☕ Kaffeelisten | React, TypeScript, Vite, Supabase, Vercel | Production-oriented PWA replacing paper-based coffee tracking at the ITC1 campus. Features company/member management, admin dashboard, and automated monthly PDF + Excel reports via Vercel Cron. Built at the B4Y3RW4LD Hackathon. Pitch deck. |
| 🗑️ Waste Sorting Assistant | YOLO, OpenCV, Python | Real-time waste bin detection system trained on 400+ manually labeled images and integrated with OpenCV for live recognition. |
| 🚢 Titanic ML Platform | React, FastAPI, Supabase, Docker | Containerized ML web application combining a React frontend, FastAPI backend, database integration, and CI/CD workflow. |
Open to jobs and collaborations · fareselbermawy@gmail.com