I contributed to a production credit-scoring model integrated into a bank’s credit-decision workflow. I prepared data, engineered features, developed gradient-boosting models, evaluated performance, and implemented a prediction API.
My projects include sentiment-driven financial forecasting and research on Implicit Causal SHAP, a counterfactual method for machine-learning explanations. I evaluated plausibility and attribution stability on credit-risk and wine-quality datasets.
At Abay Bank S.C., I provide technical and operational support for banking applications and investigate incidents affecting business operations. My machine-learning toolkit includes Python, PyTorch, and RAG, alongside experience with REST APIs.
