KPMG — Global Health Research Platform
Full-stack development on a global health platform enabling researchers to publish and manage their research projects.
A collection of research projects, side experiments, and open-source contributions.
Full-stack development on a global health platform enabling researchers to publish and manage their research projects.
AI-powered avatar that helps hospital patients with their questions, reducing the workload on nursing staff.
Production-quality machine learning pipeline for stock market prediction with ensemble models, SHAP explainability, and vectorbt backtesting.
Local AI chat application with custom agent personas, knowledge base upload, and web search — all running privately on your machine.
Master's thesis project: real-time player tracking, body orientation estimation, and scanning behavior analysis for football matches using state-of-the-art computer vision.
Event-driven trading platform with real-time market data streaming and automated execution via microservices architecture.
Predicting Acute Kidney Injury stages in ICU patients with septic shock using Decision Trees, XGBoost, and SHAP explainability on the AmsterdamUMC database.
Deep learning pipeline for 3D thoracic organ segmentation from CT scans, comparing ENet, CustomNet, and nnU-Net architectures.
Classifying 8 gym exercises from smartphone sensor data using LSTM networks, Random Forest, and signal processing techniques.
N-gram language models with smoothing and text generation, plus a neural transition-based dependency parser built in PyTorch.
Optimizing Robobo robot behavior for obstacle avoidance, item foraging, and item retrieval using CMA-ES evolutionary strategies in simulation and hardware.
Comparing Island Model evolutionary strategies with fitness-sharing for evolving neural network controllers in the EvoMan game framework.
BERT-based multilingual classifier achieving 94% accuracy on 10,000+ form entries, reducing manual processing time by 80%.
Bachelor's thesis: developing a robust BEV dataset for autonomous driving by combining drone footage, object detection, and semantic segmentation.
Investigating how knowledge distillation from GPT-2 to DistilGPT-2 amplifies gender bias in occupation-gender associations.
Computer vision system that converts sketches and logos into photorealistic neon sign visualizations with automated price estimation for a real client.