Graduate AI/ML Engineer focused on computer vision and reliable evaluation.
MSc Artificial Intelligence (Distinction) · First-Class BSc Computing.
- I build lightweight, deployable models and keep experiments leakage-safe (patient-level CV).
- I care about explainability (Grad-CAM++), tidy repos, and reproducible configs.
Core tools: Python · PyTorch · scikit-learn · OpenCV · pandas/NumPy · Jupyter · Google Colab · Git/GitHub · Java · MySQL
- GB Ultrasound (9-class) — patient-level 5-fold CV, class-imbalance mitigation, Grad-CAM++.
Repo: https://github.com/bogomil-iliev/gb-ultrasound-multiclass-cv - QC-first Ultrasound Classifier — blur/contrast/noise gates, CLAHE/NL-Means, mixed precision.
Repo: https://github.com/bogomil-iliev/ultrasound-qc-classifier - Brain MRI U-Net — medical image segmentation (Dice/IoU).
Repo: https://github.com/bogomil-iliev/brain-mri-unet-segmentation - Vehicle Rental (Java/MySQL) — RBAC + conflict-free scheduling, DAO persistence.
Repo: https://github.com/bogomil-iliev/vehicle-rental-system-java - MURA Hand (DenseNet) — study-level pipeline with per-study aggregation.
Repo: https://github.com/bogomil-iliev/mura-hand-densenet-study - House Prices — CRISP-DM workflow, baseline vs tree models (RMSE/MAE/R²).
Repo: https://github.com/bogomil-iliev/bolton-house-prices-crispdm-r - Gold Prices — Gold price forecasting in INR using Multivariate Linear Regression vs Decision Tree (quantile transform). Repo: https://github.com/bogomil-iliev/gold-price-ml/
Contact: · https://linkedin.com/in/bogomil-iliev