Rob Kras
Building the data platform foundation at Rabobank — CI/CD, infrastructure-as-code, reliable data pipelines, and observability.
I'm a DevOps and Data Engineer at Rabobank, building and operating the data platform foundation that supports analytics and machine learning workloads across the bank.
My work bridges platform engineering and data engineering — from CI/CD and infrastructure-as-code to reliable data pipelines and observability — making complex systems reproducible, scalable, and trustworthy.
I hold an MSc in Computer Science (Data Science & AI) from Leiden University and bring strong foundations in Python, SQL, and production ML workflows, with hands-on research experience in modern AI architectures.
Rabobank — Data Management & Data Products, W&R Markets and Treasury Tribe
A selection of projects showcasing my expertise in data engineering, machine learning, and cloud infrastructure.
Investigated cross-modal sound symbolism in Vision-Language Models (LLaMA 3.2-11B, Molmo 7B) via probabilistic image-text alignment and phonetic congruence analysis. Integrated SAM2 for visual grounding. Revealed a 61% bias toward pairing curved images with sonorant pseudowords.
Feature engineering, K-Folds, and ensemble methods. Discovered that simpler algorithms (KNN) can outperform complex ensembles when properly optimized. Rank 5 / 2,529.
Regression techniques with domain knowledge and SHAP for feature importance. Achieved rank 37 / 3,935. Discovered and exploited data leakage for near-perfect score.
Binary classification for financial risk assessment. Feature engineering combining credit metrics, debt-to-income ratios, and payment history. Ensemble of XGBoost, LightGBM, and CatBoost. Rank 172 / 3,724.
Predicting song tempo from audio features. Combined signal processing with machine learning, leveraging music theory for feature engineering. Rank 131 / 2,581.
Ensemble methods predicting accident risk. Created temporal and weather interaction features. Containerized training pipeline using Docker. Rank 313 / 4,082.
Customer response prediction in 7 days. YAML configuration management for rapid prototyping. Rank 576 / 3,367.