Full Stack Software Engineer (ML model development for credit risk scoring), LumiFinance — Sydney, Australia (Mar 2022–Present)
Built and deployed predictive machine-learning workflows to generate credit-risk scores from application data for downstream decisioning. Enabled model development using Python-based ML tooling and iterative evaluation to reduce processing time and improve accuracy. Designed the ML components to support production use cases where outputs are used as automated risk labels for applications. • Used Python and scikit-learn to develop and train a credit-risk scoring model • Implemented scoring logic to produce structured risk outputs per application • Tuned the pipeline to cut processing time by over 50% • Improved assessment accuracy for more reliable credit risk labeling