Mety Ayele Debebe

I've built fraud detection and transaction-risk systems using Python, scikit-learn, SHAP, and SMOTE, achieving 96% accuracy and 1.00 fraud-class precision with a Random Forest model. I also designed an explainable credit-risk scoring pipeline for 95,000+ transaction records, balancing predictive performance with Basel II auditability requirements.

During my software development internship at Kifiya Financial Technology, I supported requirements analysis, system design, testing, deployment, QA, and validation in Agile workflows. I'm currently pursuing an MSc in Cyber Security at IU Berlin and expanding my practical security skills through labs and certifications.