At ABN AMRO, I build and monitor IRB credit-risk models across PD, LGD, and in-default portfolios. I've owned annual monitoring for five models, defended reports to the Model Acceptance Group, and supported bank-wide implementation of recalibrated recovery rates.
I developed a probabilistic corporate early-warning system using financial, market, and macroeconomic signals, and carried a General Lending LGD recalibration from data requirements through business delivery.
I've also driven production tooling: rebuilding LGD monitoring on Delta Lake during a platform migration and spearheading a Python/PySpark package for automated model-performance testing as the team moved from SAS to Databricks.
Before banking, I applied Bayesian inference and deep learning to gravitational-wave research at Nikhef, Utrecht University, and Cardiff University. I built a neural-network surrogate roughly 100× faster than a reference simulator, modernised legacy scientific software, and contributed to GW150914 research.
